research methods final

bias

opinions or feelings about a specifc group of people or issue

scientific bias

tendency (conscious or subconscious) to favor some results over others

selection bias

were the participants clearly defined?

assignment bias

were groups significantly different as a result of assignment process?

performance bias

were participants aware of what group they were in?

detection bias

were those individuals measuring change aware of what group the participant was assigned?

attrition bias

did participants drop out of the study before completion of data collection?

reporting bias

authors more likely to publish positive findings

publication bias

positive findins are more likely to be published by journals

responsible conduct of research

ethical guidelines for how scientific research should be conducted and dissemated

institutional review board

reviews whether the research is ethical and to protect the participants of research as weel as the researchers

minimal risk

procedures that people regularly encounter during ordinary life (non-invasive, existing documents, anonymous surveys)

more than minimal risk

procedures that can cause stress, pain, injury, or death (vulnerable populations, invasive procedures, physical or emotional discomfort)

basic and applied research

to provide new information or to apply the knowledge in some way to change a behavior

laboratory and field research

lab is more controlled and easier to manipulate variables, field is on location

experimental research

random participant conditions, researchers control/manipulate conditions

quasi-experimental research

evaluate interventions, participants not randomized conditions

non-experimental research

no group comparison, no randomization, no measurement before comparison

independent variable

variable that is manipulated

specific trtmnt/condition/factor that is intentionally manipulated/changed/selected to see if it causes an effect

dependent variable

variable studied for change

ultimate outcome/behavior/respons that is measured and observed by researcher to determine if any changes occurred

continuous variables

range of mathematical properties

categorical variables

range of non-mathematical properties, but of categories (young, old)

internal validity

degree to which the relationship btwn independent and dependent variable is observed w/out influence of extraneous variables

external validity

generalizability of research results to other participants and settings

population validity

characteristics of participants chosen for the study are too narrow or specific, meaning results cannot be generalized to the broader real-world population

ecological validity

artificial environment of the study or participants' awareness that they are being observed causes them to behave differently than they would in a natural, everyday setting

effect size

measures the magnitude of treatment effect, independent of sample size (notated with "d"), tells how much difference a trtmnt makes

hawthorne effect

participants alter behavior simply bc they know they're being observed

observer/experimenter expectancy effect

researcher's expectations or beliefs about study's findings subconsciously influence participants' behaviors or the interpretation of results

john henry effect

experimental bias where control group, aware they are being compared to an experimental group, works harder to outperform the experimental group

halo effect

initial positive impression of a person/produce/brand based on a single characteristic, influences judgment of other unrelated traits in a positive direction

nocebo effect

describes negative outcomes that occur bc a person expects to experience them, even if the treatment or intervention is inert

social desirability bias

tendenct of respondents to answer questions in a way that will be viewed favorably by others, often presenting themselves in a positive light

implementation science

research addresses the complex, real-world variables in practice settings by investigating barriers to and solutions for delivery of sustainable, effective protocols that will maximize positive outcomes for a large number of consumers

dissemination

intentional and systematic spread of information about evidence-based interventions to specific target audiences; aim is to increase knowledge, awareness, and understanding of the intervention and its potential benefits

fidelity

degree of which an intervention is implemented as intended by its developers or prescribed in a protoco.

feasibility

extent to which and intervention can be successfully implemented or acrrried out withing a particular setting or context

penetration

measures the extent to which an EBI is integrated or saturated within a service setting and its various subsystems, often expressed as a ratio

filtered evidence

not new data, but a review of existing data

scoping review

type of literature review that aims to identify and map the existing evidence of a particular topic; explores broad topics, identifies research gaps, clarifies concepts

systematic review

research paper that collects all possible studies related to a topic; reviews quality of the papers, analyzes results

critical appraised topic

rapid review particularly useful for informing research and practice, faster than a systematic review used in EPB to assist decision-making

normal distribution

distribution or bell curve that describes a continuous probability distribution

homogenity of variance

assumption in statistical tests

parametric

interval or ratio data, large sample size, normal distribution, homogeneity of variance

nonparametic

small sample size, not normal distribution, don't assume homogeneity of variance

sampling error

natural, accidental difference btwn small group you study and the bigger group you want to help (sample size to population)

qualitative

exploring concepts, experiences, meanings through non-numerical data

understanding "why" and "how"

quantitative

measuring variables and testing hypotheses using numerical data

focuses on "how many/much" or "to what extent"

randomized control trial

experimental study design that minimizes bias by randomization, blinding, and a control group; used to assess effectiveness of intervention; highest level of experimental evidence

correlation study

important for understanding rltnshps btwn variables and identifying associations, not highest level of evidence bc does not establish causality

correlation

is there a relationship btwn x and y, and how strong?

regression

what is the relationship btwn x and y?

survey

used to collect data; questionnaires or interview questions

single-subject design

intense study of a person or small group of individuals

multiple baseline design

staggered introduction, demonstrates causation, can baseline across settings or behaviors

single-subject design

alternating treatment design

used to rapidly compare the effects of 2+ distinct interventions on a single target behavior

qualitative research approaches

ethnography

grounded thepry

phenomenology

ethnography

participant observation/interviews, flexible and exploratory, explanation from viewpoint of participant, cultural immersion over time

grounded theory

used to allow theories to emerge based on the data, coding, or categorization of data

phenomenology

homogeneous participants, semi-structured interview, broad questions, identify themes

quantitative research methods

case study

focus groups

participation action

discourse analysis

critical discourse analysis

case study

qualitative or quantitative, one case (person or agency)

focus group

group data via interview method facilitated by researcher, 6-10 strangers with shared features or interest, non-threatening enviro, planned discussion

participation action research

collaborative approach to research that emphasizes community involvement, social change, and shared learning; conducted with ppl, not on them

discourse analysis

focuses on how lang is used in text, conversation, media, and other forms of communication; how meaning is constructed through lang, how ppl use lang in diff contexts, and structure/patterns of communication

ctricital discourse analysis

focuses on specific branch of discourse that focuses on why lnag works and who it benefits, used to reinforce or challenge social inequalities, rooted in critical theory and is more politically/socially engaged

triangulation in qualitative research

strat to enhance credibility, trustworthiness, and validity of research findings by using multiple methods, data sources, theories, or investigators to study a single phenomenon

investigator triangulation

multiple researchers analyzing the same data independently, then comparing and discussing interpretations to reduce indiv biases

data triangulation

multiple data sources to examine the same phenomenon, like conducting interviews, surveys, and observations

methodological triangulation

multiple research methods to study the same phenomenon (combining interviews with participant observation or use both qual/quant methods to analyze the data)

theory triangulation

multiple theoretical frameworks to analyze data, enriching the interp and identifying potential gaps/limitations in existing theories

convergent design or mixed-method research

collect both quantitative and qualitative at same time, analyze separately, merge results

explanatory sequential design of mixed-method research

collect quantitative BEFORE qualitative

exploratory sequential design of mixed-method research

collect qualitative BEFORE quantitative

validity

accuracy and meaningfulness of measurements

face validity

whether test appears to measure what it's supposed to

content validity

extent to which the test adequately samples the content area being measured

construct validity

how well a test measures the theoretical construct or concept it's designed to assess

criterion-related validity

how will the test relates to other relevant measures or outcomes (concurrent and predictive)

concurrent validity

how well a test relates to a criterion measure administered at the same time

predictive validity

how well the test predicts future performance on a related criterion

reliability

consistency and reproducibility of measurements

test-retest reliability

consistency of results when the same test is administered at diff times to the same indiv

inter-rater reliability

consistency of results when diff examiners administer or score the same assessment

internal consistency

degree to which diff times within a test are measuring the same underlying construct

sensitivity

rate at which a test correctly identifies indiv who HAS disorder

specificity

rate at which a test correctly identifies indiv who DOESN'T have disorder

regression to the mean

if an initial measurement is exceptionally high or low (extreme), subsequent measurements are likely to be closer to the average, even if no real change or intervention has occured