SLP 510 Midterm

5 factors that contribute to the life of a theory

1. accountability

2. explanatory reference

3. testability

4. predictability

5. parsimonious

accountability

accounts for most of the data

explanatory reference

explanation must offer good grounds for believing that the phenomena would occur under the specific conditions

testability

verifiable and falsifiable

predict

predict phenomena beyond those for which the theory was originally designed

parsimonious

adopts the fewest/simplest set of assumptions

rationalism

using reason and logic to seek universal truths; rely on deductive reasoning

empiricism

observation is basis for understanding; knowledge is gained through experience; rely on inductive reasoning

research

systematic investigation; collecting and analyzing data for a purpose; experimentation aimed at the discovery and interpretation of results, revision of accepted theories or laws, or practical applications of such new or revised theories or laws

3 goals of research

1. control

2. understand

3. predict

5 major characteristics of the scientific approach

empirical referent, public/objective, repeatable, self-correcting, falsifiable

primary source

original source of research findings

secondary source

describes, explains, or interprets information contained in primary sources

tertiary source

various sources used to provide a broad overview of a topic

describe the scientific method

descriptive research

describes the current status of a phenomena; used to generate a hypothesis

analytical research

tests a hypothesis

qualitative research

systematically describes the current status of a phenomena; observations of what others do or say; focus on meaning and gaining a deeper understanding; a subjective research approach; focus on the way people interpret and make sense of their experiences

quantitative research

systematic process in which data is numerically quantified

null hypothesis

H0; default hypothesis; hypothesis that the investigator wants to reject; states that there is no differece

alternative hypothesis

H1; hypothesis attempted to be supported by investigational findings; states that differences exist

independent variable

what is being manipulated; experimental variable

dependent variable

effect of manipulation; outcome variable

"P" in PICOTS

population, patient, problem

"I" in PICOTS

intervention, issue

"C" in PICOTS

comparison, alternative therapeutic strategy

"O" in PICOTS

outcome

"T" in PICOTS

time frame

"S" in PICOTS

setting

phenomenology

subjective experiences; we can only know how things appear and never how they really are

advantages of qualitative resesarch

Focus on subjective aspects; Avoids biases inherent in quantitative approaches; In-depth examination of something; Useful for building new theories/exploring new areas of research; Dynamic and flexible; Naturally occurring environment; Fewer restrictions; Smaller sample size; Respondents provide data in their own words

disadvantages of qualitative research

Not easily generalizable Not easily replicated Lots of data that is difficult to analyze Time consuming Biases from researcher/respondent Subjective data interpretation Dependent upon researcher's skill level Small sample size Unable to establish causality

sampling techniques

convenience snowball purposeful critical case homogeneous criterion quota extreme cases typical cases maximum variation conforming/disconforming

convenience sampling

readily available participants

snowball sampling

referrals from other participants

purposeful sampling

based on emerging findings during the study

critical case sampling

participants likely to have the greatest impact

homogeneous sampling

participants with something in common

criterion sampling

participants who meet certain criteria

extreme case sampling

most extreme participants are chosen

typical case sampling

illustrate what is typical or average

maximum variation sampling

wide range of variation in participants

conforming/disconforming sampling

support or disagree with emerging explanation of findings

benefit of using open-ended questions

require more than one-word or few-word responses; encourage full, meaningful answers

3 advantages of a focus group

1. synergism

2. snowballing

3. stimulation

synergism

similar individuals are more likely to provide richer insight and foster new ideas

snowballing

one individual's comments trigger a chain reaction from other participants

stimulation

as discussion begins, level of excitement endures and individuals want to participate

open coding

data is initially placed into broad categories; subcategories are identified

axial coding

systematically examine coded data to discover relationships within and across categories; identify main themes (larger categories)

types of qualitative designs

narratives case study phenomenological study grounded theory ethonography

narrative

collecting and assembling stories from personal experiences of an individual or small group

case study

detailed analysis of 1+ individuals, groups, or institutions

case study advantages

detailed analysis

case study disadvantages

small sample size; poor generalizability; sampling bias

phenomenological study

study of individuals' lived experiences

grounded theory

developing a new theory from a corpus of data acquired by the observer

grounded theory advantages

can generate the theory during the research instead of having preconceived notions

grounded theory disadvantages

must not come with preconceived notions; must be flexible

ethnography

study of group culture

ethnography disadvantages

time consuming, requires patience

phases of SSD

baseline or withdrawal: no intervention intervention intervention 2

retrospective

use data that already exists (outcome is known); ex: medical chart review; looking back in time; less costly; shorter duration; reliance on memory/quality of information

prospective

follows participants over time; more costly; longer duration; attrition; less vulnerable to bias

case report

only one participant; very detailed description of an individual; uses descriptive statistics

case series

more than one participants who share something in common; describes characteristics or similarities; identify new disease/syndrome; describes disease progression; uses descriptive statistics

case control

case matched with a control; case is identified and a control participant is selected from same source population (must be similar to case); beneficial for rare disease/condition; can calculate risk and rate ratios

SSD

establish causal relationships in one or few subjects; series of observations over time

cohort

follows participants over time; can be closed or open

cross sectional

measures participants from representative sample at one point in time

RCT

randomly allocates participants to receive experimental condition or control condition; true experimental design characterized by manipulation, randomization, and control

randomization

each participant has an equal chance of being assigned to any group

advantages of randomization

assignment is independent of judgement bias; known and unknown variables are more likely to be balanced between groups; the larger the size, the more balanced the groups will be

control group advantages

rules out the influence of potential confounding variables; helps attribute "new" treatment to differences between experimental and control groups

single blinding

patients OR investigators are blind to their treatment assignment

double blinding

patients AND investigators are blind to their treatment assignment

triple blind

patients, investigators administering treatment, and investigators measuring response to treatment are blind to treatment assignment

what is the difference between experimental and quasi-experimental design?

randomization is used in experimental

mixed methods definition

combines qualitative and quantitative methods; uses induction, deduction, and abduction reasoning; different approaches provide different viewpoints

5 purposes of using a mixed methods design

1. triangulation

2. complementary

3. expansion

4. initiation

5. development

triangulation

both methods are used to study the same concept; focus on similarities and corroboration of results

complementary

one method is dominant and the other is used to enhance/clarify findings; focus on elaboration

expansion

different methods are used for different study components; expand breadth and range of research

initiation

discover contradictions leading to reframing of research question

development

findings from one method are used to help inform findings from the other method

equivalent vs dominant

dominant: one method is more dominant than the other, putting more emphasis on that type of design

equivalent: both methods are equally a part of the study

sequential exploratory vs sequential explanatory

sequential exploratory: qualitative first followed by quantitative; purpose is to explore relationships

sequential explanatory: quantitative first followed by qualitative; purpose is to explain relationships

systematic review

a systematic approach to reviewing, critically appraising, and synthesizing literature; clearly stated objective; pre-determined eligibility criteria; systematic search; assessment of validity of findings; systematic presentation of findings

meta-analysis

quantitative, formal, epidemiological study designed to systematically assess results of previous research to derive conclusions; pulls all data together from more than one study and analyzes it as a whole; estimate of treatment effect

forest plot

graphical display of the statistical results of a meta-analysis; shows odds ratios and confidence intervals

Type I error

false positive; incorrect rejection of the null; set alpha level

Type II error

false negative; not rejecting the null when the alternative hypothesis is supported; power of study (beta)

one-tailed testing

directional testing; predicts a difference in a particular direction; alpha is all in one tail

two-tailed testing

predicts that a difference exists, but not in a particular direction; alpha is split equally between tails

statistical significance

difference observed is above chance

clinical significance

demonstrable functional change

power

probability that you will predict if a difference exists; reduce risk of Type II error by having enough power

effect size

scale free, standardized values that provide an indication of the extent to which the value of a DV is explained by the IV

sample size

number in a study; representative portion of a population

validity

degree to which a test measures what it is supposed to measure