Home
Subjects
Explanations
Create
Study sets, textbooks, questions
Log in
Sign up
Upgrade to remove ads
Only $2.99/month
CMD 504 - Lecture 4
STUDY
Flashcards
Learn
Write
Spell
Test
PLAY
Match
Gravity
Terms in this set (66)
method section
should be a logical extension of the statement, purpose, and rationale
the author describes how the research questions were addressed and how the hypotheses were tested
components of the methods section
subject
materials
design
procedures
decision about subjects matters a lot when considering...
external validity (generalization)
________, _________, and _______ should be stated explicitly regarding subjects
who
how many
criteria
_______ and _______ could be subjects.
models and specimens
sample
a subset of the populations whose properties are studied to gain information about the whole
(share same properties from larger group)
population
a group of individuals, objects, or items from which samples are taken for measurement
sampling
measuring a small portion of something then making a general statement about the whole thing
process of selecting a number of units for a study in such a way that the units represent the larger group from which they are selected
sampling: the process of selecting units from a _____________ so that by studying the ________, you can ________ your results to the population from which the units are chosen.
population of interest
sample
generalize
understanding sampling requires understanding its relationship to ___________.
external validity
why do we need sampling/ what is it used for?
sampling makes it possible to study large, heterogenous populations
economy, speed, accuracy
major problem with samping
if sampling is biased, not representative, or too small, the conclusion may not be valid or reliable
Economy
Smaller size of people is cheaper but if a sample is too small it may not generalize
Accuracy
If you study too many people in a complex study, error will occur
(less likely to make an error with smaller amount of people)
Speed
Will be done quicker with fewer people
probability samples
simple random
systematic random
stratified
multi-stage cluster
non-probability samples
convenience
snowball
quota
theoretical
the random sample
allows a known probability that an individual unit will be chosen
what is used in the lottery and raffles
if you want to select 10 players randomly from a population of 100, write their names, fold them up, mix them thoroughly, then pick 10.
simple random sampling
objective - to select n units of N such that NCn has an equal chance of being selected
(online survey, phone survey)
-Generate random numbers & select one person per random number
what is the procedure for simple random sampling
get a list or "sampling frame"
this can be difficult because it cannot systematically exclude anyone
systematic random sampling
select a random number, which we will know as k
gather a list of people or observe a flow of people
choose every kth person, but make sure that there is no rhythm
ex. every fifth person being a senior
stratified random sampling
separate your population into groups or "strata"
take either a simple random sample or stratified random sample in each subgroup
can do proportionate or disproportionate stratified random sampling
advantages of stratified random sampling
ensures better representation of key subgroups in the population
improved statistical precision due to subgroup homogeneity
when using the same sampling fraction within the strata
proportionate stratified random sampling
when using different sampling fractions in each strata
disproportionate stratified random sampling
multi-stage cluster sample
divide the population into groups called clusters or get a list of "clusters" if already existing
randomly select clusters from the list
randomly sample people within the selected clusters (five counties from all the counties of NY)
when is the multi-stage cluster sample useful?
when sampling a population that is spread across a wide geographic range
non-probability sampling
sampling that does not involve random selection
two broad types: accidental and purposive
convenience (accidental) sampling
find people that are easy to find
ex. person to person street interview, asking people to volunteer, finding patients within the speech clinic that would fit your study
purposive sampling
you sample with a purpose in mind
you are likely to get the opinions of your target population, but you are likely to overweigh subgroups in your population that are more readily accessible
quota sampling
determine what the population looks like in terms of specific qualities
create "quotas" based on those qualities
select people for each quota
you sample until you achieve a specific number of sampled units for each quota
proportional quota sampling example
If 80% of customers are females and 20% are males:
sample 80 females and 20 males for N = 100
disproportional quota sampling
If 80% of customers are females and 20% are males:
sample 50 females and 50 males for N = 100
advantage of quota sampling
allows for oversampling of smaller subgroups
disadvantage of quota sampling
it is likely to be more biased than stratified random sampling because it often depends on who comes along when
snowball sampling
find a few people that are relevant to your topic
ask them to refer you to more people
especially useful when you are trying to read populations that inaccessible or hard to read
the problem is low external validity
sample size
total number of participants included in the study
qualitative researchers seek _________.
saturation - want to have a deep understanding, number is not important
quantitative researchers seek ________.
statistical validity
factors that affect sample size
the purpose of the study
variability
research design (within subjects you need half the amount of people)
concern about generalizability
sample size is important because...
contributes to the power of a research design
increasing a sample size can improve...
generalizability because it can make for a more representative sample
what three things do you need to calculate sample size?
statistical test
confidence interval
effect size
generalizability of research findings
1. randomization - selecting random sample of people to be included in the study, can enhance generalizability because it could make for a more representative sample
2. alternative techniques - stratified random sampling, cluster sampling, multistage sampling
Random Assignment
2 groups and each subject has equal change of randomly going into each group (reinforces internal validity of study)
Random Selection
Select subjects randomly from population (increases generalizability... aka external validity)
systematic replication
improves external validity through direct replication and generalization across populations, settings, or other variables
value: affirms research results
protection of subjects/ participants
1. respect for persons
2. beneficence
3. justice
____________ serves as the foundation for the conduct of research in the U.S.
The Belmont Report (1979)
materials
identification of materials used to measure or generate the independent or dependent variables in the study
describes the materials, measures, equipment, or stimuli used in the experiment (includes testing instruments, technical equipment, books, or images)
measurement
the assignment of numerals to objects or events according to the rules
what is an important goal in CMD research in terms of measurement?
the measure of speech, language, and hearing variables with a clear and practical set of rules
nominal level of measurement
distinguished by name only. no agreed upon order of data
example: gender, color of eyes, yes/ no responses, name
counting the frequency of occurrence is the only mathematical operation possible
ordinal level of measurement
ordinal level is similar to nominal in that it is distinguished by name, but there is an ordering scheme
cannot measure the difference between two pieces of data
example (list of top ten popular cities to live, letter grades, severity of aphasia)
interval level of measurement
interval level can be ordered
differences between data are meaningful and can be measured
data at this level lacks a zero starting point, so ratio comparisons are meaningless
(temperature, standard scores on behavioral tests)
ratio level of measurement
data at the ratio level possesses all of the features of interval + a zero value
(height, amount of salary, suttering frequency)... most accurate
reliability of measurement
refers to the degree to which we can depend on a measure
two types of errors that may influence the reliability of the measurement process
1. systematic: recurs consistently with every repeated measure
2. unsystematic: unpredictable
estimating measurement reliability
1. stability of measurement (test re-test)
2. equivalence of measurement
(alternate or parallel forms are equivalent)
3. internal consistency of measurement
(split half method)
observer reliability
to account for measurement errors that may emanate from the observer who measures
interobserver agreement coefficients
derived from measurements made by two or more observers measuring the same event
intraobserver agreement coefficient
compare the measurements made by one observer at two different times
behavioral instruments
1. standardized instruments: look for evidence of reliability and validity
2. nonstandardized instruments: look at the manner in which the instrument was constructed in order to judge its adequacy
procedure
description of what you had participants do, how you collected data, and the order in which the steps occurred
there are factors that need to be controlled in making measurements in CMD research that have to do with the procedure aspect of the study
factors to be controlled in terms of procedures
1. test environment
2. subject instructions
3. observer bias
4. appropriateness of measurements
data analysis
a research paper typically includes this subsection in Methods section
describes how the data will be organized, summarized, and assessed
all software and hardware employed in data collection, storage, and analysis should be clearly described
Other sets by this creator
Fluency Exam 2
22 terms
Dysphagia Midterm
46 terms
Dysphagia Quiz 1
21 terms
Fluency Test 1
26 terms
Other Quizlet sets
COSM 1000 Infection Control
148 terms
EAS457
20 terms
Astronomy II: Post Midterm Content (Chapters 9, 10…
47 terms
Male reproductive and nephrology P&R
54 terms