Research Methods ch. 1-5
** R O1 X Y1 O2 R O3 Y1 O4 R O5 X Y2 O6 R O7 Y2 O8
"Factorial or Multivariable Design"; examine effects that manipulation of at least 2 IV(simultaneously at diff levels) has on DV -impact that each IV has on DV= main effect -DV impacted by interaction of IV= interaction effect
** R X O1 R O2
"The Posttest-Only Design"; experimental, randomized, control & exp. group, only given posttest
** R O1 X O2 R O3 O4
"The Pretest-Posttest Design"; experimental -randomized, control & exp. group, randomized, pre& posttest
** R O1 X O2 R O3 O4 R X O5 R O6
"The Solomon Four-Group Design"; 4 groups randomized -assesses effect of treatment in relation to control group, determine effect of pretest, explain interaction btwn treatment conditions & pretest
Experimental research
-Designed to establish cause (IV) & effect (DV) relationships; study conditions are controlled; eliminates variables where there's no relationship
*Belmont report
-Ethical principles & guidelines for protection of human subjects research *addresses 3 ethical principles: justice, respect for persons, beneficence
types of IRB categories
-Exempt: must notify IRB & complete forms to exempt project from full IRB review -Expedited: for projects involving no more than minimal risk to subjects -Full board: if greater than minimal risk present
the research process (IDC-MIR-CAD) (1/2)
-Identify problem; Develop hypothesis/ research questions; Conduct literature review; Make research design; Identify tools you'll need
Modern era of research:
-Informed consent -Right to privacy & confidentiality -Institutional Review Boards (IRB)- approve research involving ppl -Researcher training (need proof of training)
internal vs. external validity
-Internal: shows whether IV was the only cause of change in DV -External: shows extent results are applicable to real world -Both: need to balance them
Belmont report: 3 ethical principles
-Justice: treat ppl equally -Respect for persons: acknowledge autonomy & protect those w/ diminished autonomy -Beneficence: respect ppl's decisions, protect from harm, & secure their well-being
extraneous threats to internal validity
-Maturation, testing, instrumentation, sequence effect -experimental mortality/attrition, contamination
*examples of theories
-PRECEDE-PROCEDE, Transtheoretical Model of Change, Social Cognitive Theory, Diffusion of Innovations Theory
examples of applied research
-Polling ppl on health matters/ public opinion about smoking -Social impact of peer groups, social media, or public policy; Health behavior change; Needs assessment
clinical trials
-RCT= inmedical & pharmaceutical studies; test efficacy of intervention against controlled condition -strict inclusion & exclusion criterion of who can participate -4 phases (w/ diff. objets, design factors, duration, ppl, sample size)
*Functions of IRB (RDPM)
-Review all research efforts involving human/animal subjects -Develop policies on research -Provide education to investigators -Maintain records (up to 3-7 years) after research finished
*PICO acronym
-Richardson et al (1995); serves as framework to develop BEST q & guide researcher in right direction -P: patient/problem I: intervention C: comparison (when used) O: outcome
peer-reviewed literature
-Using key words, search data-bases relevant to topic; data-bases relevant to health sciences: Cochrane Library, PubMed, SUMSearch
government documents
-accepted as forms of evidence; most commonly found online -Ex: National Center for Health Statistics (NCHS), National Health and Nutrition Examination Survey (NHANES), National Vital Statistics System (NVSS)
*Institutional review board
-approves studies; looks at methodologies/whether beneficial for public or not -DHHS & FDA require any institution involved in research to have IRB -IRB process may differ slightly btwn institutions
statement of the problem good & bad examples
-bad: a health promotion program for better eating habits -good: Does a health promotion program for junior high school students increase nutritional food selection in the school cafeteria?
sequence effect
-can be confounding b/c person's performance in later condition could be result of role in previous condition of study -Ex. exercise physiology study
instrumentation effects
-changes of instrument/ measuring device used to test effect of DV -& observers or raters can change rating system -Ex. procedures for recording data (scanner sheets) should remain same
confidentiality vs. anonymity
-confidentiality: ppl can see who participated in study but not shown results -anonymity: can see others' results w/out names (ppl referred to as #s) -both: anonymous
research proposal
-document that outlines study; written by researcher- justifies why study should be done -written in FUTURE tense
*Helsinki Declaration
-est. 1964; updated 2000 by World Medical Association (WMA) *focus: biomedical research of human subjects by physicians (strict guidelines on what can/can't be done) *another response to Nuremberg
What are some ways that confidentiality and privacy can be broken?
-exposing results not in peer-reviewed journal; records stolen -subject is relative/friend of research team who has limited access to data
informed consent
-gains ppl's permission to have research done on them; protects participants -provide clear description of project & what's involved, time, compensation, any benefits/ potential risks -state ppl's right to leave (at no cost), ability to ask q's & to sign it
testing effect
-in pre-post tests, pre-testing may sensitize subjects when taking test 2nd time (subjects may act diff. if no pretest measures taken)
dissemination of research
-investigator= responsible to disseminate findings (whether results answered hypothesis/q's or not
Internal validity
-level of control has DIRECT effect on internal v; greater control, greater internal v. -difficult to control outside lab envir.
Annotated Bibliography
-list of citations, books, other empirical evidence used for research topic -citation: short paragraph or up to 8 bullet points -narrative: describes key points in study
hypothesis
-logical supposition/ suggested answer to a problem (based on previous studies) -can be prediction of outcome; must be TESTABLE *imp. to support it w/ peer-reviewed literature
experimental mortality/attrition
-losing subjects during study; subjects CAN be replaced if done equally in both groups
Vulnerable groups: kids
-need consent; ppl giving consent need to be given same info that adult part. would be given -children/ minors need same rights as adult participant
dependent variable (DV)
-outcome variable; criterion by which results of experiment are judged -dependent on manipulation of indep. variable
*null hypothesis
-particular outcome isn't true; no relationship btwn variables; p ≤ 0.5 -states that minor diff. between variables can occur b/c of CHANCE errors, not sign. difference -reveals existence of factor that caused statistical diff.
true experimental designs
-posttest-only control group -pretest-posttest control group -solomon 4-group -factoral/ multivariable design
X O
-pre-experimental design; treatment to 1 group, no control -observations on treatment in posttest; WEAKEST design
X O ----- O
-pre-experimental; experimental & control group, not random -each group given posttest
O1 X O2
-pre-experimental; pretest, experiment, posttest; 1 group tested -no control, not random
primary vs. secondary data
-primary: data collected yourself (interviews, trials) -secondary: articles w/ already done research (census)
What is Research?
-process of collecting & analyzing info to increase understanding of phenomena; imp. to communicate understanding to others *used in conjunction w/ theories; it's how we justify what we do
Critical Review of Literature
-provide framework for research q, hypothesis, & setting of problem -developed thru systematic review of lit, reference books, & abstracts thru searching catalogs using "key words"
Theory
-provide ways to: conceptualize (put into perspective issue dealing with), organize, integrate, & classify data *in heath field, based around BEHAVIOR change
stating significance & justification of problem
-provides backing for why study conducted -states potential benefits -researcher(s) must justify study- provide evidence in form of *review of literature*
Critical Review of Literature- purpose
-provides understanding in theory; defines problem -reviews procedure & instruments used; originates NEW ideas (not work already done) -uses recommendations from existing research -critiques relevant studies to apply well-founded components to own study while avoiding pitfalls
quantitative vs. qualitative research
-quantitative: gather & analyze numerical data -qualitative: gather & analyze descriptive data; use inductive reasoning for analysis (surveys, interviews; gather ppl's opinions; time consuming)
O1O2O3O4 X O5O6O7O8
-quasi; no control, multiple observations before & after intervention; not random -many extraneous variables controlled
X1O X0O X1O X0O
-quasi; not random, no control -treatment given & observed; treatment not administered but "event" is observed - process repeated
O1 X O2 ----------------- O3 O4
-quasi; not randomized, control & experimental group -both groups given pre & posttest
methodology
-research proposal should include detailed methods section- how study conducted -include selection of ppl, instrumentation, how data, time frame
Vulnerable groups: pregnant women
-research similar to children concerning criterion -has to be a risk-benefit ratio conducted to ensure no harm to mother or child
Applied Research
-solve practical, real life problems -theoretical concepts tested in real situations (ex. using dummies in heard impact car crash- volvo)
independent variable (IV)
-stands alone, not changed by other variables; "cause" in cause & effect -manipulated independently from other variables
Annotated Bibliography- key points included
-study size (population/sample), purpose, methods (instrumentation), results, conclusions -reference= approx. 150-200 words
maturation effect
-subjects mature/change over time; during long treatment, subjects get hungry, tired, board -hard to avoid in longitudinal studies
*why would reading or looking in textbooks be a good place to start in research?
-textbook=secondary source; have bibliography (provides reader w/ primary sources)
Comprising a problem statement: components needed to consider/state
1) Limitations 2) Delimitations 3) Assumptions 4) definition of terms
2 cases that influenced changes/improvements on human subjects research
1) Nuremberg Medical Trials: testing WW2 Jews w/out permission; led to Nuremberg Code 2) Tuskegee study of untreated Syphilis; tested blacks for syphilis when penicillin available
types of designs
1) pre-experimental: lowest level evidence, no control, little/no control of EV; not randomized 2) quasi-exp.: 2nd highest, not all EV controlled, not randomized 3) experimental: highest level, control & exp. groups, randomized
how to eliminate testing effect:
1) use comparison group exposed to DV but doesn't get pretest 2) increase time length btwn pre & post-test
*selecting research problem (1/2)
1. **Interest -most imp. 2. operability- need time & resources 3. scope- can't solve all health problems nor lack info for adequate results 4. theoretical & practical values- benefit health field 5. health paradigm- model employed by researcher 6. values of researcher- researcher- aware that values may prejudice research
*The scientific approach (1/2)
1. goal= find solution to problem 2. Research develops generalizations, principles, theories- predict future events 3. R is based on observations or empirical evidence 4. Accurate observation & description of phenomenon
*selecting research problem (2/2)
7. research methodology- appropriate hypothesis & methods 8. reactivity- method of data collec. examined for reactivity 9. unit of analysis- note whether research applicable for indiv. or pop. 10. time frame- limited for students doing research 11. budget-ensures feasible proposal
*referencing
=approx. 150-200 words -APA, MLA, AMA -choosing style depends on journal you submit work to
Basic research
Aim to expand knowledge by forming, evaluating, or expanding a THEORY.
suppressing data for publication
Don't get results wanted but include it; esp. imp. to avoid for drug research
tools to catalog & organize research
EndNote, Zotero, RefWorks
hypothesis example
Health promotion programs for junior high schools will increase the selection of nutritional foods in the school cafeteria.
In an experiment looking at weight loss drugs, what would be the IV and the DV?
IV= drug, DV= weight loss
In an experiment measuring temperature and solubility, what would the IV and DV be?
IV= temperature, DV= solubility
*PICO example
In infants with bronchiolitis (patient), does surfactant (intervention) compared to placebo (comparison) lead to a reduced length of stay (outcome)?
ethical issues arising in research (2/2)
Inadequate data storage, Misleading authorship, Suppressing data for publication, Sponsored research, Therapeutic research, Copyright info, protection of subjects
ethical issues arising in research (1/2)
Plagiarism, Research question/topic selection, Fabricating data, Recruiting research participants & Faulty data collection (sample bias)
the research process (IDC-MIR-CAD-WS) (2/2)
Recruit participants; Conduct study; Analyze data; Develop conclusions based on results; Write Report; Spread report
*The scientific approach (2/2)
Research... 5. gathers new/existing for new purpose 6. requires expertise-existing knowledge of issue 7. strives to be objective & logical 8. find answers to difficult q's -requires patience & disappointment
null hypothesis examples
There will be NO differences in the selection of nutritional foods in the school cafeteria between those exposed to the health promotion program and those not exposed to it.
what is not appropriate to use when doing research?
Wikipedia, youtube, people magazine, etc.
Design symbols
X= Independent variable (1st treatment) X0= treatment not given Y= Independent variable (2nd treatment) O= process of observation (2 levels of O's means there's control group R= random assignment to groups --= no random assignment
Types of Research
basic, applied, experimental, quasi-experimental, surveys
Delimitations
boundaries made BY researcher; what won't be included
Limitations
boundaries of problem est. by factors/ ppl OTHER than researcher
Assumptions
conditions taken for granted & w/out which research would be impossible
Vulnerable groups: elderly
considered vulnerable when memory loss illnesses apparent, and/or they're institutionalized
order of deliver
each group gets treatment done at same time; if exp. requires same subjects to be exposed to 2+ treatments, error may occur due to ORDER treatments are present
external validity threats
ecological validity
extraneous variable
effects DV in a way not related to purpose of research; need to control this before est. cause & effect relationship
*Nuremberg Code
est. 10 guidelines for human subjects research (voluntary consent, exp. should yield fruitful results, should avoid all physical/mental suffering, degree risk< degree of benefit)
population validity
generalizing study pop. to general population
how to get started w/ research
get understanding of topic (read related secondary sources/textbooks)
copyright information
give credit where credit is due; get permission to use/publish copyright materials
statement of the problem
gives focus & direction to project (statement or q); clear & concise; written in education level that reflects participants -needs to state specific topic & pop.
therapeutic research
human experimentation intended to benefit the subjects on whom it's performed (drug testing)
randomization of situational variables
if observer/facilitator, etc. used as part of study, groups randomly assigned to that person
misleading authorship
if part of study & didn't complete what you've done /didn't include author after contributing
fabricating data
if you don't like exp. results, manipulate data way you want
how have these codes, declarations, and reports influenced research on human subjects today?
influenced whole structure of how we do research
blinding
participants don't know whether receive treatment or placebo -double blind= person doing treatment/person recording results don't know if in exp. or control group
constancy of conditions
ppl in exp. & control groups exposed to identical situations (except for differing conditions of IV)
randomized matching
ppl matched on as many EV's as possible, then randomly assigned to group -Ex. 2 partners match based on age, then randomly assigned to control or ex. group
random assignment
ppl randomly assigned to control or exp. group; purely by change
contamination
researcher has previous knowledge of ppl in study; cause researcher to treat indv. w/in group differently -blinding eliminates this
Definition of terms
researcher must define each term that relates to project
quasi-experimental designs
subjects not randomly assigned; usually carried out on intact groups due to availability
what constitutes a research project?
unanswered & unsolved question, concern, query, & statement of inquiry.
mixed methods research
using both qual. & quantitative data -Ex. Protective sex at Stetson; design surveys or construct problems w/protective sex (qualitative) -then collect numerical data
