Intro to Research
About this set
Created by:
dourada Plus on January 31, 2011
Subjects:
Description:
Intro to research mid-term review questions
Classes:
Log in to favorite or report as inappropriate.
Order by
74 terms
Spanish | English |
|---|---|
 pure | → abstract and general → concerned with a brand new theory/new paradigm → often claimed to be done for its own sake → ex/ Einstein (theory of relativity); Galaleo |
| applied | → designed to answer a practical question → based on previous knowledge → ex/ Newton: why did apple fall down?; does chiropractic work? |
| experimental | → evaluates impact of manipulating one variable on another → variable's effect on an outcome (manipulate variable) → control group → ex/ comparison studies |
| descriptive | → first report of something new/unique experience with something (case/treatment) already known (watched over time) → looks at current reality to gain insight on trends → ex/ case studies! |
| clinical | → done in "real" world setting, where control often quite difficult to do (mimic reality in a lab; little control) → may be observational, applied, descriptive, or experimental → ex/ randomized clinical trial |
| laboratory | → done in artificial/unreal/simplified/controlled type of setting → ex/ lab rat studies |
| randomized clinical trial | ex: of experimental applied research |
| tradition | → it's the way it's always been done (paradigm) → ex/ cultural tradition, family tradition → rehabilitative sciences/practices are steeped in tradition → ex/ chiropractic techniques → problem: because it's always been done the same way, some things are never evaluated for validity/ compared to better alternatives |
| Authority | → because the doctor says so! → ex/ names: Dr..., ...PhD, Peg Phillips, etc.) → ex/ books: Bible, Torah, Koran, Green Books → can lead to dogmatism (i.e. Galen's views on anatomy/physiology...not questioned); until Versalius, da Vinci, Harvey questioned and laid foundations for relying on one's own observations |
| Trial and error | → keep trying until something works → no systematic approach, haphazard → expensive (time and money) → process ends when satisfactory solution is found (may not go further to find best/perfect solution) → ex/ Thomas Edison (discovered filament for electric lights by accident); Paul Ehrlich (discovered chemotherapy/chemical formulations) |
| Logical reasoning | → it stands to reason, thus it must be true → credited to ancient Greeks -two kinds: deductive and Inductive |
| Deductive | - starts with general premise/proposition and goes to specific - conclusion in form of syllogism: kind of logical argument in which one proposition (the conclusion) is inferred from two others (the premises) of a certain form - premises must be true - can generate hypotheses - cannot be new source of knowledge - ex/ chiropractic: premise of universal intelligence |
| Inductive | - starts with specific and generalizes - ex/ case study approach |
| Quantitative Paradigm methods | -scientific method 1.Theory- developed a priori (before the fact) 2.Subject Selection- randomly; reasonable sample size needed to derive significant results (calculated before implementation) 3.Measurement - instrumentation (i.e. tools, tests, interview protocols) 4. Manipulation - generally of only one variable (controlled) 5.Control - essential to eliminate/reduce extraneous variables (to accurately elucidate cause/effect relationship) |
| Quantitative paradigm assumptions | 1. reality: single, objective: single, objective reality and it is based on our ability to measure natural phenomena (aka empiricism) - goal is to understand and reliably predict and/or control reality 2.investigator and subject: independent of one another (possible to study something unobtrusively) 3. generalizable : desirable and possible --goal of research to develop generalizable characteristics of reality 4. causes and effects: causal- can be determined/ differentiated from each other - concern: whether relationship between independent (cause) and dependent (effect) variables can be clearly shown 5. research is value-free (i.e. use of control groups for reference eliminates investigator biases in interpreting significance of findings) |
| Qualitative paradigm methods | 1.Theory- allowed to develop as investigator proceeds - one begins with an idea of what concepts may be important- process often open-ended 2.Subject Selection-non-random, targeted - considered non-problematic since statistical analysis is not done 3.Measurement - interviews/narratives of/by subjects- records feelings, perceptions rather than objective facts 4.Manipulation - no manipulation is done; simply reported/described 5.Control - no control - setting is natural, may change with each individual and interactions |
| Qualitative paradigm assumptions | 1. reality: multiple, constructed- realities (i.e. there are several versions of reality) - truth about anything is relative - ex/ patient responses to same chiropractor: cool vs. weird 2. investigator and subject -dependent (process of inquiry affects both) - ex/ chiropractor participating in study of professional demeanor likely to change behavior during observation 3. Generalizable: Situation Specific--knowledge is time- and context-dependent (i.e. idiographic/ pertinent only to particular case under study) 4.causes from effects: Non-causal--impossible to distinguish (i.e. it's better to describe than impart meaning that allows/suggests control or predictability) 5. research is value-bound (i.e. reflected in types of questions asked, how constructs are defined/measured) - all inquiry influenced by values of inquirer and funding resources |
| Single System Paradigm | Assumptions → same as quantitative, except: - treatment effectiveness is subject- and setting-dependent → subject-centered ▪ Single-System methods: 1. Theory - a priori (same as quantitative) 2. Subject Selection - targeted, purposive 3. Measurement - objective, standardized 4. Manipulation - same as quantitative 5. Control - same as quantitative |
| hypothesis | a testable prediction, often implied by a theory |
| variable | measurable indicators of a characteristic, whose value may change in individual or whose value may differ among members of a population (ability to measure a variable , is dependent on the ability to define it) types: continuous, discrete, dichotommous, independant, dependent, confounding, or random |
| statistic | numerical characteristic of a sample from a population (roman letters denote) number that correlates, organizes and summarizes a group of recorded variables-collecting, classifying, and analyzing data |
| parameter | a numerical characteristic of a population (greek letters denote) - ex/ mean, standard deviation, variance, correlation, proportion |
| association | an observed relationship between two or more variables |
| operationalization | aka operationalized concepts to make operational; to define an abstract concept in such a that it can be practially measured, in statistics this requires assigning numbers to attributes (like visit frequency, patient satisfaction etc) so that they may be analyzed (this process of quantifying/counting is actually the process operationalization) |
| paradigm | one that serves as a pattern or model form or model for how things work |
| algorithm | a precise rule (or set of rules) specifying how to solve some problem -established and well defined step by step method used to achieve desired mathematical results -ex: PV+, PV- |
| spurious association | a casual relation is apparent in two variables, when in fact none exists because a third, hidden variable causes both |
| construct | an abstract or general idea inferred or derived from specific instances, not real |
| n=1 study | single system study like quantitative variables-objective, theory-priori focuses on one target population |
| Incidence vs prevalence of subluxation | prevalence measure of total number of cases in population-how widespread the "disease" is incidence- rate of occurrence of new cases-risk of contracting the "disease" |
| content validity vs criterion-related validity | content-extent to which an indicator(s) measures full domain of interest -how effect a test is at actually measuring what it claims to criterion-extent to which one test predict or agree with the results of gold standard |
| sensitivity vs specificity of a screening | sensitivity-ability to identify those who truely have the "disease" (=A/A+C) specificity-ability to identify those who truely do not have the "disease" (=D/B+D) |
| reliability vs validity of tests | reliability-repeatable validity-is it measuring what it is intended to measure (general valid tests are reliable, but reliable doesn't always indicate vailidity) |
| Quantitative vs Qualitative | theory-a priori vs open ended subjects-random vs targeted measurement-instrument vs narrative(subjective) manipulation-one variable vs all variable |
| ratio vs proportion | ratio-comparison of 2 statistical values(5men:15women) proportions-expression of equivalent ratios(5men:20population) |
| ratio | the relative magnitudes of two quantities (usually expressed as a quotient) |
| proportion | the quotient obtained when the magnitude of a part is divided by the magnitude of the whole |
| PV+ vs PV- | positive vs negative predictive value PV+ a screening tests ability to correcting identify those with the "disease" PV- a screening tests ability to correctly identify those who do not have "disease" prevalence of "disease" important -low prev PV- more likely to be accurate -as prev increases PV+ will also increase PV+= A/A+B PV-=D/C+D |
| normal vs negative distribution | normal-parabolic curve even on both side mean=median=mode Negative-skewed to the right, mean<median<mode Positive-skewed to the left mode<median<mean |
| mean | the arithmetic average of a distribution, obtained by adding the scores and then dividing by the number of scores |
| median | The middle number in a set of numbers that are listed in order |
| mode | the most frequent value of a random variable |
| Intra vs Inter-examiner validity | Intra-extent to which same examiner will produce same results Inter- extent to which multiple examiners will get same results |
| True/False Positive/Negative | TP-those who test positive for X that have X (box A) TN-those who test negative for X who don't have X (box D) FP-those who test positive for X that don't have X (box B) FN-those who test negative for X that have X (box C) |
| Randomization and blinding | attempt to reduce bias of test by gathering an untargeted population, so as to recreate a semblance of the general population, and by disallowing information to possibly bias an examiners observation or patient's experience |
| tradition vs authority knowledge | authoritative knowledge can become traditional over time -the way we have always done it -because so and so said so |
| trial and error vs scientific method | -haphazard manipulation until success -logic, hypothesis, and systematic variation to control and prove/disprove |
| descriptive and inferential statistics | -descrp/parametric to describe what's going on in our data, organizes and summarizes observations (mean, median mode, standard deviation) -to make inferences from our data, draw conclusions about conditions |
| N vs n | N represents entire population n represents population studied |
| continuous vs discrete | -observed and reported on a linear scale, continuous units -limited, discontinuous range of observable data, measurable in whole units |
| independant vs dependant variables | -intervention thought to influence event -correlated direct effect on outcomes |
| systematic and random measurement error | discordance between what was measured and what was actually occuring sys- arising out of a faulted system rand-happenstance |
| nominal/classification | uses discrete data only, and these data are utilized as category labels and have no logical order -can only do frequency and mode |
| ordinal/rank order | uses discrete data, but this data is at least ranked -giving numbers inherent worth -can analyze count, mean, mode |
| interval data | continuous data with equal intervals between each, no zero point -mean, mode, count |
| ratio scale | continuous data with equal intervals and a set zero point -all analysis applicable |
| Continuous variable | -e.g. joint ROM,strength, distance, weight, chronological time, height -value anywhere along a continuum within a defiend range |
| discrete variable | -e.g. heart beat -measured in whole units -nominal scale data is discrete by definition |
| dichotomous/binary variable | -e.g. yes/no, male/female, case/not case - discrete that can only have certain values |
| Independent/manipulable/predictor variable | -e.g. drug use, surgical intervention, chiropractic adjustment -characteristic or experimental intervention thought to influence a particular event/manifestation -not all independent variable are manipulable |
| Dependent/outcome variable | -e.g. recovery, addiction, death, complications -resulting outcome/event/manifestation |
| confounding variable | -e.g. lung cancer risk from radon confounded by smoking -e.g. chronic cough from exposure to workplace SO2 confounded by smoking -one which distorts the degree to which an independent variable affects a dependent variable, may be due to an association with one or both variables |
| random variable | -one whose value is determined by chance -applicable to studies in which participants are picked at random |
| nominal/classification scale | -discrete data -e.g. sex nationality, blood type, clinical diagnosis -only frequency and mode |
| ordinal/rank order data | -discrete but ordered -numbers indicate rank- are qualitative -e.g. pain index, grading and staging of neoplasms, likert rating scale -count, mode, and median only |
| interval data | -continuous (parametric) -e.g. calendar yr, IQ, temperature -doesn't represent absolute magnitude |
| ratio scale | -units with equal intervals measured from a true zero -quantitative -e.g. distance, age, time, weight, blood pressure, ROM |
| face validity | -most basic and least rigorous e.g. ROM (goniometer), length (ruler), strength (dynamometer), sensation (pin prick), gait, and balance obvious and non-controversial |
| content validity | -e.g. theoretical domain or universe of behavioral variables -extent to which an indicator measures full domain of interest -questionaires, examinations, inventories, interviews -e.g. choosing NBCE questions |
| criterion-related validity | -measure results of target test to predict or agree with gold standard -most objective validity -TB skin test vs X-ray vs culture -to what extent is leg length inequality an indicator of LBP |
| construct validity | -degree to which one measure correlates with other measures of the same abstract concept -e.g IQ as a measure of intelligence, functional status, quality of life perception |
| internal validity | -when differences between treatment and control group outcomes are attributable only too independent variable -adjustment vs emotional support |
| external validity | -degree to which research findings can be applied to general population -vaccine efficacy |
| predictive validity | -ability to predict outcome -e.g. advanced ed --> financial success or X-ray abnormal mass --> malignancy |
First Time Here?
Welcome to Quizlet, a fun, free place to study. Try these flashcards, find others to study, or make your own.