Social Work Research
Epistemology: how we know what we know, way of understanding
Authority: understanding through role models, or someone who has more experience or education.
Experience: understanding the ways of life through experience
Media: world wide, accessible, and different perspectives of information
Tradition: way a family or community practices or engages in knowledge, it's consistent, and part of one's identity and cultural heritage
TROUT: Tentative, Replicable, Observable, Unbiased, Transparent
Phases of Research Process
Placebo effect: a "fake" treatment. Anything that seems to be a "real" medicine treatment but isn't.
Recognizing flaws in unscientific sources: inaccurate observation, overgeneralization, selective observation, ex post facto hypothosizing
Evidence based practice: is considering values and expectations of clients and involving them as informed participants in the decision making process.
CIAO: Client characteristics, Intervention being considered, Alternative interventions, Outcome
Positivism: the philosophy of quantitative research
quantitative research: objective observation, controlled manipulation, reliability through numbers. A philosophical system that holds that every rationally justifiable assertion is capable of logical or mathematical proof and that rejects metaphysics and theism. Emphasizes the production of precise and generalizable statistical findings. deeper meanings of particular human experiences, and generate theoretically richer observations.
Constructivism: the philosophy that says all knowledge exists because humans say so
Subjective nature of knowledge, knowledge is built on agreement
Mixed methods research: stand alone research design in which a single study collects both quantitative and qualitative data. Integrates both sources of data at one or more stages of the research process to improve the understanding of the phenomenon being invested
- formulate the problem 2. design the study 3. collect data 4. process the data 5. analyze the data 6. interpret the findings 7. write up and share results
Kinds of studies
- Exploratory study: an early study looking at natural phenomenon and figuring out what it is
- Descriptive study: looks at things more specifically and try to describe it
- Explanatory study: why does something happen
- Evaluation: experiments what the outcome will be
- Constructing measures: asking specific questions based on symptoms
NASW Code of Ethics in Research
voluntary participation & informed consent, no harm to participants, anonymity & confidentiality, deceiving participants, analysis & reporting, weighing benefits and costs
Internal Review Board
Group of professionals who review scientific proposals to ensure human protection. consider their influence early on, respect for person, beneficence, and justice
Cultural competence: Access to the population, learn as much as you can, cultural humility, hire people form their community, take care of potential barriers
Independent variable: intervention - "causer"
Dependent variable: symptoms/problem you're trying to address "causee"
mediating variable: mechanism to affect relationship by independent and dependent variables
Moderating variable: mediating variable that strengthens a relationship
control variable: moderating variale actively studied
Conceptualization: how you conceptually define your variable
Operalization: how it is measured
NOIR: Nominal: named, Ordinal: ranked variables
Lit Reviews: describes problems, what has happened in the past, why you're doing it. to give consent of the outline of the paper
Systematic error: consistently gives you poor data or information
Random error: using cumbersome, complex, or boring procedures
Reliability: inter-observer, test-retest, parallel forms, internal consistency
Validity: criterion Related, predictive, know groups, construct, convergent, discriminate
Sampling: generalized or generalizable
sample error: difference between sample size and the population
Population: group of potential units that could be sampled
elements: the individual parts
parameter: description that delineates who can be sampled
sampling frame: list from which elements can be selected
probability sampling: randomly selecting
Systematic random sampling: start at a random number and follow specific ordinal after
stratified sampling: group populations in homogenous groups before you sample and then randomly sample
Cluster sampling: randomly sampling from the cluster
Non probability sampling: not as effective, convient
deviance case sampling: based on bell curves
quota sampling: have a predetermined quota of what the sample consists of and pick from there
snowball sampling: finding one person that brings in more people
intensity sampling: looking outside the norm
theoretical sampling: sample from population until you do not get any infromation
Survey: satisfaction with services, teacher evaluations, marketing hook, describing the population, needs of groups.
correlation: 2 variables that are related in terms of frequency and intensity. Correlation does not cause causation. One event must precede the other. Eliminate competing variables.
external validity: the ability to generalize findings
Internal validity: the level of confidence that you can say the independent variable changed the dependent variable
Threats to internal validity
History: external events are going to happen that affect the study
statistical regression: going away from the norm
maturation: changes in body and age
ambiguity of timing: not sure what causes what due to timing
testing: using the same test can alter findings
Instrument changes: changing instruments during the study.
measurement bias: garbage in garbage out
selection bias: researcher has role of who goes into each group
Researcher reactivity: the way the data is presented by the researcher
random sample & random selection
placebo effect
Attrition: drop out rate
Single subject design: test intervention/control group. Well defined dependent variable. Independent variable can be taken away. measure baseline -- intervention -- baseline
triangulation: reinforcing data with more data
Coding: data entry, use NOIR variables,
Procedure for selecting statistics
- questions to address
- find the scales and variables that gives you your answer
- identify nature of variables
- draw diagrams for the questions
- determine parametric or non-parametric statistics
- choose your test/make choice
Qualitative Research: knowledge is inherently a subjective experience, filtered through humans, basis of communication, never objective, verified through consensus, understood through paradigms
constructivism: all knowledge is related to experience
Ethnography: knowledge is built through accurate and detailed descriptions of the way people live
Feminist: knowledge is best framed from experiences of those who biologically differ from the majority
Moral Activism: knowledge is contextually political
Critical Theory: knowledge comes from the framework of culture, society, and political structure
Cultural Studies: knowledge encompasses all aspects of cultures highly self reflective
Queer Theory: Knowledge is built from framework of sex, gender, race, and culture to the end of de-simplifying words used to describe experiences
Participatory Action: knowledge comes from those studied they have ultimate say in how knowledge is used
Grounded Theory: knowledge comes from identifying patterns as best possible independent of previous knowledge or context
Phenomenology: knowledge is based on experience as understood through states of consciousness
Secondary data: data someone else collected, can be anywhere or anything
Manifest: summarizing what is said
Latent: summarizing what is meant
The Four Walls
Wall 1: antecedent
Wall 2: physical setting
Wall 3: social structure
Wall 4: intention for audience
Logic Model
Inputs: everything needed to make the agency function
Process: services provided
Outputs: the number or statistics of who is there
Outcomes: if there are any improvements in symptoms
Approach to Program Evaluation
Needs Assessment: surveys, rates, social indicators
Process evaluation: staff and client surveys
Goal attainment: quasi or experimental designs
Cost efficiency: cost benefit analysis