Advanced Research methods

Midterm

Analysis

exploring thoroughly how a paper is built and its contents

Assessment

having a judgement about the relative applicability of a paper for understanding, therapy, educating

critique

assessing strength of arguments and supporting evidence

research

intense seeking for knowledge (beyond Google)

literacy

knowledge and ability to understand ideas and applications

research literacy

knowledge of and ability to understand research

Pros of research literacy

confidence in knowledge

better reading comprehension

less worry when writing/working with patients

greater academic success

Cons of research literacy

more time spent reading

annoyance with simple conversations

challenging your basic assumptions of own knowledge

bias

unreasonable and undeserved preference

unspoken assumption

prejudice

inflexible and irrational attitudes/opinions

schizophrenia people are dangerous

ideological dominance

one theory and only that

quality/rigor

stats and execution

to be a researcher

follow set of rigorous methods/principles

use careful measurements (q’s)

rule out alternative explanations for ideas

account for bias

internal and external validity

what research does

shows us what we know

shows us what we do not know

research approach

overarching plans/procedures for conducting a knowledge gather activity (experiment)

includes research design, methods, problems

research design

specific framework for conducting study

inquiries(quantitative, qualitative, mixed methods)

research method

specific tools to use within said framework

self-report

research problems

should be well defined

no confounds

quantitative approach

numbers driven

tests

what degree, how often, what conditions variables relate or interact

qualitative approach

text driven (linguistics)

what are lived experiences of such group

interviews/open ended q

labor intensive

themes created based on results

data that numbers can not find

mixed methods

both quantitative/qualitative

triangle relationship philosophy, method, design

all work together to construct a valid, coherent, and rigorous study

worldview

general philosophy understanding of how world operates

4 worldviews

postpositivist

constructivist

transformative

prgamatic

postpositivism

encompasses positivism/post positivism focused on scientific method and certainty in results

well defined/operationalized terms

facts/truths are created by data

looks at relationship between variables

objectivity, bias at minimum

research aims to clarify or abandon ideas/theories

data/evidence allows for knowledge

constructivists

subjective, unique to the individual

based on lived experiences

history, context needed to make sense

open ended q to find human constructs of meaning

interaction is cause of meaning

transformative

political, social change

feminists, empowerment, oppression, domination, suppression, alienation, critical theory

voice for marginalized groups

inequality happens based on diversity variables (gender, age, sexual orientation, disability)

participants come up with q

pragamatic

actions, situation, consequences

looks at anything/everything,

no commitment of specific philosophies

ultimate freedom in research

nothing is universal

truth is what works at that specific moment

what and how for getting a conclusion

quantitative research design

verifying the truth of an idea, theory, concept using data/number

survey research

experiments

correlational

longitudinal

cross-sectional

meta-analysis

qualitative research design

exploring/describing experiences of individuals/groups

narrative research

phenomenological research

ethnography

case study

narrative research (qualitative)

long form interview for one persons lived experience

phenomenological research(qualitative)

described lived experience of a group for one phenomenon

grounded theory (qualitative)

from interviews develop theory of process/action/interaction in the phenomenon

ethnography(qualitative)

explore shared patterns of behavior/language/action of a culture or group in a natural setting over long period of time

jane goodall

case study (qualitative)

one person one phenomenon, one period of time

freud- anna O

Mixed Methods design

integrating quantitative and qualitative

convergence of data types for collecting and analysis

explanatory sequential

exploratory sequential

explanatory sequential

quantitative then qualitative

harder to do because it takes a long time

exploratory sequential

qualitative then confirm with quantitative

subject may be tired after interview to answer questions for data

how to collect information for research methods

standardized tests/measures

naturalistic observations

interviews

literature review

behavioral observations

topics for research lit review

effectiveness

prevention

risk/prognosis

assessment

description

framework for topic types PICO

patient/group/population

intervention

comparison/control group

outcome measures

Can vs should

can study be completed

should the study be completed

can?

is there funding?

time?

participants available?

tools/measurements?

should?

add to literature?

add to the profession?

purpose of study serve?

challenge a theory?

too simple/complex?

risk of harm?

purpose of literature review

share results of similar studies

informs reader of trends

establishes benchmarks

identifies solutions to the problem/topic

sets tone for rest of article

types of quantitative lit reviews

integrating- mix or related/unrelated literatures

criticizing- previous literature/authors

bridge building- between concepts

identifying- central issues in a field of study (broad)

types of qualitative lit reviews

ethnographies

grounded theory

case studies

phenomenological studies

all begin with subject assumptions to help reader

types of sources for lit reviews

primary

secondary

Primary lit review source

original studies

dissertations

technical reports

meta analysis (sometimes)

secondary lit review sources

meta analysis

research reviews (APA= bias

lit reviews (case studies, highly bias, make arguments)

practice guidelines (organization like APA)

organize literature

topic

studies in 3 sub categories

subcategories of lit explained/labels key words

Key words/terms for lit review

define when first presented

operational definitions, commonly accepted in field

do not use common slang

Lit review structure

intro of review

review first topic

review second topic

review third topic

summary

Internal validity

finding the underlying causes of a phenomenon

concerned with accurately finding cause and affect variables

randomized controlled trial (RCT)

gold standard of research

allows for internal validity

randomization

comparison/controlled groups

spurious relationship

false relationship between two variables because both are caused by a third

randomization

helps control for extraneous variables

assign sample to control/comparison group

correlation

when two or more variables are related, relationship could be true or false

“r”

threats to internal validity

history of participant life

maturation of person (age)

mortality of participants may leave before study is over

testing- participants response (demand characteristics)

instrument decay- may not be the best for the study

regression- performance drops off the more you do it

placebo effect- change behavior because they expect to change

manage threats to internal validity

control groups (RCT, watch for selection bias)

experimental control

statistical power high (bigger sample)

quasi-experimental design (pre/post test)

experimental designs

random assignment and comparison groups

pretesting helps small studies

pretesting helps dissect for mortality/other issues

Solomon four group

controls for pretesting and participants become aware of study

control and experimental group receive pretest

control and experimental group receive none

Within subjects design (repeated measures)

participants take part in different trials

with and without experimental condition

drawbacks of within subjects

fatigue of participants

ordering effects; condition or not first?

not good for long lasting treatments (medications)

Efficacy

high experimental control

important for clinical research

should say straightforward how much an intervention worked

Effectiveness

less on control on threats to internal validity

more on how treatments could work in different settings/groups

strengths of experimental designs

rule out confounds

determine causality

greater internal validity chances

weaknesses of experimental designs

not ethically study everything (no experiment on child abuse)

artificial in nature hard for ecological validity

lacking in detail due to operationalization of terms

external validity

how well results generalize to other settings

ecological validity

populaitons

large groups of people for a certain phenomenon sa

sample

smaller subset of a general population

sampling

helps control for external validity

probability

non-probability

probability sampling

randomly picking people from known population to be in sample

picked from list of knowns (study burnout in therapists, pick psychologists)

simple random sample (probability)

all potential participants are assigned a random number

systematic sampling (probability)

starting at a random point on a list, then choosing every 3rd person

stratified sampling (probability)

divide sampling frame into different groups (ethnicity)then use simple/systematic

multistage cluster sampling (probability)

sample in progressively narrower stages

look at states, school districts, schools, then pick the people

non probability sampling

intentionally picking people from the known population

poor external validity/worse ecological validity

convenience sampling (availability)

choose from near by groups

not representative of populations

cost effective

recruited by ads= biasq

quota sampling/stratification

making sure enough participants from different groups are in the sample

randomness leaves study

purposive/judgement sampling

researcher chooses best fit

qualitative research

snowballing

participants are asked to bring in anyone they know may be interested

common for survey research

sample sizes

large= more accuracy, costs more money/time

small= less accuracy, less cost/time

sample size determined

relative population size

previous research

statistical test that will be performed in data analysis (quantitative)

Power analysis

determine optimal sample size

ANOVA, Chi-Squared, T-Tests, effect size, significance level(p=.05)

always conduct prior to the stidy implementation

Instrumentation

tools used to collect data

they need to be valid

content of instruments

do items assess what the designers meant them to assess

Construct of instruments

do the items measure the construct at hand

Predictive/concurrent instruments

do scores of one measure correlate with a similar measure

not many have good discrimination

Instrument reliability

does the meausere consistently produce results expected

Internal consistency of instruments

how often/similar items identify an underlying construct

more items on test greater chance for consistency to go down

test-retest reliabilty

degree to which a test produces similarly results across time

Data analysis

explore response bias

look at the descriptive analysis (mean, SD, ranges, demographic data)

inferential data(correlations significant?)

effect size significant?

confidence interval

Cohens d effect size signficance

.2= small

.5= medium

.8= large

confidence intevcal CI

95%= 95/100 times observed score will fall in that range