Final Exam Study Guide Flashcards by Libby Sarra

what is a population

group who you are interested in learning something about

How well did you know this?

Not at all

Perfectly

group who you are interested in learning something about

Population

How well did you know this?

Not at all

Perfectly

what is a sample

small group of people from TP included in the actual study (study participants)

How well did you know this?

Not at all

Perfectly

small group of people from TP included in the actual study (study participants)

sample

How well did you know this?

Not at all

Perfectly

where can you find the sample information

methods section

How well did you know this?

Not at all

Perfectly

what is the sample size and what is an appropriate size

how many people being studied; red flag <20 people
Small leads to inconclusive results or random chance dominates outcome

How well did you know this?

Not at all

Perfectly

what is a representative sample

reflects characteristics of TP
Provides generalizability
Similar to the TP

How well did you know this?

Not at all

Perfectly

reflects characteristics of TP
Provides generalizability
Similar to the TP

representative sample

How well did you know this?

Not at all

Perfectly

what is a power analysis

determines if the research manuscript considered sample size when designing their study

How well did you know this?

Not at all

Perfectly

determines if the research manuscript considered sample size when designing their study

power analysis

How well did you know this?

Not at all

Perfectly

what is generalizability influenced by

sampling bias & inc/exc criteria

How well did you know this?

Not at all

Perfectly

what has an impact on generalizability

sampling bias, sample error, small sample size

How well did you know this?

Not at all

Perfectly

what is generalizability closely tied to

external validity

How well did you know this?

Not at all

Perfectly

what is external validity

extent to which the findings can be applied to broader populations

How well did you know this?

Not at all

Perfectly

what is a non-representative sample

only relevant to that sample & not TP
Biased samples, Over and under estimate certain pop attributes, impacts overall validity

How well did you know this?

Not at all

Perfectly

only relevant to that sample & not TP
Biased samples, Over and under estimate certain pop attributes, impacts overall validity

non representative sample

How well did you know this?

Not at all

Perfectly

How to know if sample represents TP

similar profile

How well did you know this?

Not at all

Perfectly

what is sampling

choosing sub-group to represent TP
Process used to pool the sample from the TP
How the sample was chosen

How well did you know this?

Not at all

Perfectly

what is sampling error

difference between the makeup of the study sample and that of the population of interest
Happens naturally

How well did you know this?

Not at all

Perfectly

what is sampling error influenced by

sample size & variability in TP

How well did you know this?

Not at all

Perfectly

what reduces sample error

larger sample

How well did you know this?

Not at all

Perfectly

what is sampling bias

certain members are more/less likely to be included in a sample than others

Meaningful differences between non-participants and participants

How well did you know this?

Not at all

Perfectly

Meaningful differences between non-participants and participants

sampling bias

How well did you know this?

Not at all

Perfectly

how to determine sampling bias

Look for how participants are chosen, demographics are broad, groups that were explicitly excluded?

How well did you know this?

Not at all

Perfectly

what is a consort diagram

insight into how many potential participants actually enrolled & completed the study

consort diagram

Used to record the study outcomes Indicates if intervention is effective or not Are they reliable & valid

measurements

what are measurement scales

quantify, categorize, analyze & report data

what are the measurement scales

ratio interval ordinal nominal

describe ratio

cannot have - #s, true 0, equal interval bw units, allows for math (+,-,/,x): ex duration, SPL, grip strength, height, weight

describe interval

can have - #s, no true 0, equal interval bs units, ex range of motion, BC thresh, temp, allows for math

describe ordinal

inherent ranking/ordering, #s have no value, order of finishing a race, agree on satisfaction survey, Likert scale

describe nominal

categorize/label variables, #s have no value, no order, ex 1-f 2-m, 3dead

describe binary

only allows for two values, ex below 50 & above 50 or yes no

what are the 3 sources, how do you fix them

instrument (calibration), person (training/unfamiliar w/ instrument), variable (inherent instability - think BP variability; take multiple measures)

what are the types of errors

systemic & random

describe systemic errors

occurs in 1 direction & constant magnitude; graph is straight line

describe random errors

unpredictable, due to chance; graph is random

concerned with degree of random error in measurement process

reliability

Asking if measures are

dependable, reproducible, & if measured more than once is the result the same

what are the types of reliability

test retest rater - intra & inter internal consistency

describe test retest reliability

measuring repeatedly & compare results; error due to instrument

what are the rater reliabilities

inter and intra

describe intra reliabiity

1 person measuring same construct multiple times & do they get same answerd

describe inter reliability

multiple raters and calibrated to measure the same thing, do they get the same result

describe internal consistency

related to questionnaires & reliability bw set of items measuring same concepts (depression, pain, anxiety, happiness, etc.)

what are good and bad reliability coefficients

<.5 poor, .5 - .75 moderate, >.75 - 1 good Cannot be negative or below 0 Context specific (psychological, motion, ROM, physiological, chemical) - look at previous lit

A measure with perfect agreement will never demonstrate a perfect correlation

FALSE always

Perfect reliable instrument = observed measurement is true

true

what is validity

is instrument measuring what is intended to

what are the types of validity

face content construct criterion

what is face validity

instrument is testing what it is supposed to, Weakest form

face validity

what is content validity

items making up instrument adequately sample universe of content that defines the variable measured; useful w/ questionnaires & inventories Experts subjectively determine if an instrument captures all domains of a construct

content validity

what is construct validity

establishes ability of instrument to measure abstract construct & which it reflects theoretical components of the construct (QOL, depression, burnout etc)

construct validity

types of construct

convergent discriminant

describe convergent construct validity

tests that are intended to measure same thing actually does so Compare your outcome measures to established outcome measures of the same construct - should have similar results Have one test already established & you want to compare your new one

convergent construct validity

describe discriminant construct validity

tests that are measuring diff things actually do so (depression from anxiety) Take an already established outcome measure that assesses another related construct and compare your outcomes to it Scores should differe because they are measuring different constructs (discriminates between the two)

discriminant construct validity

what is criterion validity

assess ability of 1 test to predict results obtained from another (criterion) test; most practical & objective Indicates outcome of 1 instrument, target test, can be used as a substitute measure for an established gold standard criterion test

criterion validity

what are the types of criterion validity

concurrent predictive

describe concurrent criterion validity

compares less established/new instrument to gold standard measuring the same thing with 2 instruments @ the same time

describe predictive criterion validity

correlates one thing Ex: gre test scores to measure succes in grad school or correlates gpa in science to future performance

predictive crierion validity

compares less established/new instrument to gold standard measuring the same thing with 2 instruments @ the same time

concurrent criterion validity

Questions to ask in Instrumentation or Outcomes section

What are DV of the study How is each variable operationalized Potential range of values/scores for each instrument & how are they interpreted Is there evidence of instrument reliability & validity

what is the target population

group to whom you can generalize your results

how do you operationalize the target population

Inclusion-characteristics of the population you are interested in studying; characteristics they have to qualify for the study Exclusion-identify attributes that confound study findings

what is the inclusion criteria

characteristics of the population you are interested in studying; characteristics they have to qualify for the study

what is the exclusion criteria

identify attributes that confound study findings

what is the independent variable

variable being manipulated, differs between groups, typically identify levels of the independent , always described as a construct (treatment or intervention)

how to operationalize the IV

distinguish bw levels of IV & replicate same procedures in another study

what is the dependent variable

the outcome of the study/variable which you expect the see a diff bw groups

how do you operationalize the DV

describe how it is being measured (what measurement tool will be used to measure the variable & describe the measurement procedures enough detail to replicate the measurement

what is internal validity

are results caused by the thing you are testing or is it something else influencing it? Accuracy of study, Did show that IV and only IV caused results.

what is external validity

degree to which the results of study can be generalized to the larger target pop, application to other situations how you can relate the study to the outside world. Can we generalize results beyond the sample, decreases as internal increases

Which of the following can directly influence the generalizability of a study? Sampling bias Scales of measurements Inclusion and exclusion criteria All of the above Sampling bias and inclusion/exclusion criteria

Sampling bias and inclusion/exclusion criteria

When recruiting subjects, you should consider whether non-participants (those who decline to be included in the study) differ from participants (those who agree to be included in the study) in any meaningful way. If there is a meaningful difference between non-participants and participants, this may indicate: Sampling bias A strict exclusion criterion The need to stratify your sample None of the above

Sampling bias

The study sample will be described in the _________ section of a manuscript. Discussion Abstract Introduction Methods

Methods

The generalizability of a study is closely tied to: Internal validity Statistical validity The p-value Operationalization of the dependent variable External validity

External validity

A representative sample is a sample that reflects: Characteristics of the target population The researchers' ability to access the target population Sample size All of the above

Characteristics of the target population

Sampling error refers to: A respondent's adoption of a meaningless response pattern, such as "a,b,c. . .a,b,c. . a,b,c." The number of groups in the study Critique provided by an expert panel The difference between the makeup of the study sample and that of the population of interest Within group variability

The difference between the makeup of the study sample and that of the population of interest

In order to determine if the authors of a research manuscript considered sample size when designing their study, you should look for evidence of: Table 1 A power analysis A statistically significant effect Effect size calculations

a power analysis

A ___________ will provide insight into how many potential participants were actually enrolled in, and completed the study. Copyright issues Table 1 Power analysis Significant result CONSORT diagram

CONSORT diagram

A subset of individuals from the larger group of individuals we want to learn something about is called the Sampling Sample Population Independent variable

Sample

In most cases, it is impractical to study an entire population because (check all that apply): It would be cost prohibitive It would be time intensive It would be too efficient It would be difficult to identify everyone in a given population

It would be cost prohibitive It would be time intensive It would be difficult to identify everyone in a given population

A sample that is not representative of the population (check all that apply): Can underestimate certain population attributes being studied Can overestimate certain population attributes being studied Is also known as a biased sample Will impact the overall validity of a study

Can underestimate certain population attributes being studied Can overestimate certain population attributes being studied Is also known as a biased sample Will impact the overall validity of a study

Generalizability is dependent upon A representative sample The independent variable A sample that reflects the relevant variables of a population A sample that reflects the relevant characteristics of a population The dependent variable

A representative sample A sample that reflects the relevant variables of a population A sample that reflects the relevant characteristics of a population

Which of the following can impact the generalizability of study findings? Sampling error Sampling bias Small sample size All of the above

all

Measurements are an essential component of any research study.

If you cannot trust the study measurements, you cannot trust the study findings.

When critically reviewing measurements from a study, you are primarily scrutinizing the measurement's: Validity Randomization Sampling methods Reliability

validity reliability

Which of the following measurement scales would always allow you to implement mathematical operations? Ordinal Ratio Nominal Interval

ratio interval

Which of the following correctly ranks the measurement scales from most informative to least informative? Ratio, interval, ordinal, nominal Interval, nominal, ordinal, ratio Nominal, ordinal, interval, ratio Ratio, ordinal, interval, nominal

Ratio, interval, ordinal, nominal

I measured my patients’ handgrip strength using a dynamometer. I record the measurements in kilograms then classify each patient as 1=weak, 2=normal, 3=strong. I have converted my data from a(n) _____________ scale to a(n) ______________ scale. Interval, ordinal Ratio, interval Interval, nominal Ratio, ordinal

ratio ordinal

I measured my patients’ handgrip strength using a dynamometer. I record the measurements as 1=weak, 2=normal, 3=strong, then classify patients as 1=cleared for participation, 2=not cleared. I have converted my data from a(n) _____________ scale to a(n) ______________ scale. Interval, nominal Nominal, ordinal Ordinal, nominal Ratio, ordinal

ordinal nominal

Primary sources of error include: The variable The instrument The rater All of the above

all

A _______ error will ALWAYS be in the same direction and in the same magnitude. Random Rater Response variable Systematic

systematic

When determining the reliability of a measure, you are likely asking: How is the dependent variable operationalized? Are the measurements reproducible? Are the measurements dependable? If the variable is measured more than once, will you get the same result? How is the independent variable operationalized?

Are the measurements reproducible? Are the measurements dependable? If the variable is measured more than once, will you get the same result?

Acceptable intra- AND inter-rater reliability should always be established prior to data collection.

____________ reliability is concerned with the stability of data recorded by one rater across two or more trials.

intra

Raters and response variables are not considered during _______________ reliability. Internal consistency Intra-rater Inter-rater Test-retest

test retest

Which of the following is true? A measure with perfect agreement will always demonstrate a perfect correlation A perfectly correlated measure will always demonstrate perfect agreement A perfectly correlated measure will never demonstrate perfect agreement A measure with perfect agreement will never demonstrate a perfect correlation

A measure with perfect agreement will always demonstrate a perfect correlation

When an instrument is perfectly reliable: The true measurement is random error plus systematic error The observed measurement is the true measurement The observed measurement is the true measurement plus random error The true measurement is the observed measurement plus systematic error

The observed measurement is the true measurement

If you, as a clinician, review an instrument and make a subjective judgment call that the instrument measures what it intends to measure, you are determining: Concurrent validity Face validity Convergent validity Content validity

face

If, during the development phase of an instrument, a researcher asks experts (eg, researchers, clinicians, patient) to subjectively determine if an instrument captures all domains of a construct, the researcher is aiming to establish: Concurrent validity Face validity Convergent validity Content validity

content

Which of the following are classified as types of construct validity? Convergent Concurrent Predictive Discriminant

convergent discriminate

Which of the following are classified as types of criterion validity? Convergent Concurrent Predictive Discriminant

concurrent predictive

If I measured my subjects' activities of daily living using a scale in which the total score consisted of the sum of their endorsements of 20 items, but chose to report the percentage of subjects who scored below 50 and those who scored above 50, I would be converting my data from a(n) _____ scale to a(n) _____ scale. nominal, ordinal ratio, binary interval, ratio ordinal, ratio interval, ordinal

ratio binary

Authors of a research manuscript conventionally provide evidence of a scale's reliability or validity by citing a(n): Paper in which the scale was used previously Validation article Personal communication from the scale's developer Analysis conducted during the current study

validation article

You are interested in studying the effects of prolonged exposure to loud noises on hearing ability. At the end of your study, you realize that your audiogram underestimated hearing threshold by exactly 10 decibels for all study subjects. This means that your hearing ability measures are: Reliable and valid Not reliable but valid Reliable but not valid Neither reliable nor valid

reliable but not valid

You are interested in studying the effects of caffeine (low dose vs. high dose) on heart rate. For your study, you have operationalized heart rate as 1=below normal, 2=normal, 3=above normal. Which of the following mathematical operation(s) are permissible on your dependent variable? Multiplication Addition Counting None of the above

counting

You are interested in studying the effects of caffeine (low dose vs. high dose) on heart rate. For your study, you have operationalized heart rate as average number of beats per minute during a 5-minute session. Which of the following mathematical operation(s) are permissible on your dependent variable? Counting Multiplication Average Division All of these operations

all

I read a manuscript that reported developing a scale I want to use. The authors reported that the scale is valid. How do I determine if it is reliable? Find another development article for the scale Assume that the scale is reliable because it has been shown to be valid Find another article on the scale’s reliability Find another scale

Assume that the scale is reliable because it has been shown to be valid

The developers of the GRE argue that it is capable of predicting performance in graduate school, but the administration at Harvard claims that it does not because it is biased. They are arguing about: Face validity Content validity Reliability Criterion validity Discriminant validity

criterion

One version of the SF-36 quality-of-life scale has been validated for acute stroke patients. The authors of this scale noted that the population they used for validation had suffered a stroke within the last 60 days, and were more than 65 years of age. It is permissible to use this scale for: Patients who suffered a stroke within the last 60 days Patients who are more than 65 years of age Patients who suffered a stroke within the last 60 days, and are more than 65 years of age Stoke patients Patients over 65 years of age

Patients who suffered a stroke within the last 60 days, and are more than 65 years of age

The quality of a manuscript is best evaluated: Either awful or great On a continuum Black or white Either yes or no Blue or red

continuum

The authors of a research manuscript used the Neurobevioural Functional Inventory, but report no validation information for this scale. You are interested in potentially using the results of their study to modify your practice. What should you do? Search for a validation article for the Neurobevioural Functional Inventory Ignore the authors’ study because they were too stupid to include validation information Assume that the Neurobevioural Functional Inventory is valid, and go forward Validate the Neurobevioural Functional Inventory yourself

Search for a validation article for the Neurobevioural Functional Inventory

Which of the following are primary components to evidence-based practice? Clinical expertise Patient values Reliable measurements Best available evidence

Clinical expertise Patient values Best available evidence

Primary research articles collect data from: The participants Published research articles The peer-reviewers Textbooks

participants

Secondary research articles collect data from: The participants Published research articles The peer-reviewers Textbooks

published research articles

Systematic reviews and meta-analyses can be very useful for busy clinicians because the authors do all of the hard work for you including Completing a comprehensive search of the literature to find relevant articles Critically reviewing all of the relevant articles Incorporating the evidence into your patient care Offering a concise summary of the findings from the articles

Completing a comprehensive search of the literature to find relevant articles Critically reviewing all of the relevant articles Offering a concise summary of the findings from the articles

The basic structure and sections of the methodology of systematic reviews and meta-analyses include the: Search strategy Selection criteria Assessment of study quality Synthesis of the evidence

Search strategy Selection criteria Assessment of study quality Synthesis of the evidence

All systematic reviews and meta-analyses are of good quality.

how certain are they that the DV caused the IV

bias

if they can convince that manipulation of IV resulted in change in DV you get low risk of bias scores in most categories

true

Primary components to evidence-based practice

Clinical expertise, PT values, & best available evidence

Primary research methodology (randomized controlled trials, cohort studies, etc.)

Study design Participants Procedures study/main outcomes Data is collected from participants

Secondary research articles methodology (systematic & meta-analysis)

Search strategy, selection criteria, assessment of study quality, synthesis of evidence

Review of databases & sources used to identify relevant articles

search strategy

Review of the inclusion & exclusion criteria of the articles

selection criteria

Ensures high-quality studies (e.g., well designed) are weighted more in synthesis of findings

assessment of study qulity

Stage at which investigators determine if meta is appropriate

synthesis of evidence

Meta is pursued if included studies are

similar enough

Can also use _________to determine if meta is appropriate

statistical measures of heterogeneity

If I square is you can consider meta

<.6

Uses quantitative methods to summarize current status of evidence by pooling results of similar studies asking the same question

meta

what is the purpose of i squared

describes the % of variation across studies in a meta-analysis that can be attributed to heterogeneity beyond chance

what is a forest plot

graphical representation displaying pooled results of included studies

what is the 95% CI

Estimates what would likely happen if the entire pop was included in the study want the line smaller

triangles/sample size of forest plot

contribution of the study to the overall results

weight of forest plots

How much each study contributes to overall effect It is a fxn of the sample size and how wide the standard deviation was (smaller = more weight) & how they are calculated

overall large triangle at bottom of forest plots

Pooled effect size of all included studies

Solid vertical line in forest plot

null line or line of null effect No difference between intervention & control group Divides one result from the other - ratio

Each of the following would be included in the methodology of a systematic review EXCEPT: Synthesis of evidence Statistical analysis Search strategy Study quality assessment All of the above would be included in a systematic review

Statistical analysis

The fundamental methodological difference between primary research and secondary research is their __________________ . Outcome variables Target population Unit of analysis Intervention of interest

Unit of analysis

The overall quality of a meta-analysis will depend upon __________________ . The number of articles included in the analysis Its common effect The reported effect size across all studies Its methodology

methodology

An I-squared value of .90 would suggest that: The effect of the pooled data significantly favors the intervention group The effect of the pooled data significantly favors the control group A meta-analysis should not be performed Data from multiple studies can be pooled to derive one overall estimate

A meta-analysis should not be performed

The methodology of a systematic review should be explicit and structured to ensure that readers can: A. Reproduce the study B. Assess the overall quality of the study C. Determine whether the proper statistical analyses were utilized for the study D. Both A and B E. All of the above

D. Both A and B

In a forest plot, if the diamond for “overall effect” crosses the vertical line at “0” this indicates: The effect of the pooled data significantly favors the intervention group The effect of the pooled data significantly favors the control group The difference between the pooled groups is not statistically significant Significant heterogeneity

The difference between the pooled groups is not statistically significant

Which of the following would you tend to trust most? Meta-analysis of cohort studies Meta-analysis of randomized controlled trials Meta-analysis of cross-sectional studies Systematic review of case studies Systematic review of case-control studies

Meta-analysis of randomized controlled trials

A systematic review’s “target population” is operationalized by the Objective Search strategy Selection criteria Bias evaluation I-squared value

selection criteria

Which type of research is primarily focused on numerical data and statistical analysis? Qualitative research Quantitative research Both Neither

quant

Qualitative research methods include surveys with closed-ended questions. True False

false

Which approach is more likely to use participant observation and interviews for data collection? Qualitative research Quantitative research Both Neither

qual

Cross-verifying data through multiple sources or methods

triangulation

what is triangulation

use of two or more strategies to collect, interpret or analyze information, e.g., combining data from multiple sources: interviews, observations or multiple raters

Enhances the credibility of the findings Validating findings with study participants

member checking

what is member checking

Investigators verify their interpretations about the data with the subjects who provided the information

Systematically maintained set of documentation

audit trail

what is the purpose of audit trail

To ensure the research process was systematic and transparent, To check for the confirmability of the study outcomes , and To verify the consistency and dependability of the findings

accuracy and truthfulness of the findings

credibility

applicability of the study's findings to other contexts

transferability

stability and consistency of the data over time

dependatbility

degree to which the findings are shaped by the participants and not researcher bias

confirmability

similar to objectivity

confirmability

similar to reliability

dependability

similar to external validity

transferability

similar to internal validity

credibility

Transferability in qualitative research is similar to which concept in quantitative research? Internal validity Validity Reliability Generalizability

Generalizability

What does credibility in qualitative research refer to? The statistical significance of the results The size of the participant sample The potential to replicate the study The accuracy and truthfulness of the findings

The accuracy and truthfulness of the findings

Which of the following best describes member checking's contribution to research? Increases numerical data accuracy Reduces the need for data analysis Enhances the credibility of the findings Decreases the study's duration

Enhances the credibility of the findings

Credibility in qualitative research is similar to which concept in quantitative research? Generalizability Validity Reliability Internal validity

nternal validity

Triangulation in qualitative research refers to: Using three data collection methods only Limiting research to three participants for depth Focusing solely on triangular relationships Cross-verifying data through multiple sources or methods

Cross-verifying data through multiple sources or methods

What is the primary focus of qualitative research? Statistical analysis Understanding human experiences Predicting future trends Generalizing findings to large populations

Understanding human experiences

Which is NOT a purpose of conducting an audit in qualitative research? To ensure the research process was systematic and transparent To verify the consistency and dependability of the findings To assess the financial expenditures of the study To check for the confirmability of the study outcomes

To assess the financial expenditures of the study

Dependability in qualitative research is similar to which concept in quantitative research? Reliability Internal validity Validity Generalizability

Reliability

Member checking involves: Checking members' credentials in a research team Peer review of the research methodology Membership in relevant research organizations Validating findings with study participants

Validating findings with study participants

Which data collection method is commonly used in qualitative research? Surveys with closed-ended questions In-depth interviews Numerical data analysis Large-scale experiments

In-depth interviews

for qualitative studies, which of the following study features would likely indicate a good quality study? A control group External auditor Member checking Random assignment Triangulation

External auditor Member checking triangulation

Which is of the following questions would be associated with the critical review of the study sample? **Check all that apply** Is the study sample large and diverse enough to generalize study findings to the target population? Does the target population look like your patient population? Are the inclusion / exclusion criteria too strict or narrow? Did the investigators provide enough details of the intervention for a reader or researcher to replicate? Were participants recruited from multiple locations to create a diverse study sample?

Is the study sample large and diverse enough to generalize study findings to the target population? Does the target population look like your patient population? Are the inclusion / exclusion criteria too strict or narrow? Were participants recruited from multiple locations to create a diverse study sample?

Which is of the following questions would be associated with the critical review of the independent variable? Did the investigators provide enough details of the intervention for a reader or researcher to replicate? Is the study sample large and diverse enough to generalize study findings to the target population? Were participants recruited from multiple locations to create a diverse study sample? Did the investigators provide enough details for the reader to distinguish between the levels of the independent variable? Are the inclusion / exclusion criteria too strict or narrow?

Did the investigators provide enough details of the intervention for a reader or researcher to replicate? Did the investigators provide enough details for the reader to distinguish between the levels of the independent variable?

In critically reviewing a description of the target population, what are you looking for?

Inclusion/exclusion criteria External validity (can the results be generalized, and do these subjects look like my patients?) How were the subjects recruited (e.g., single center or multiple centers)

Critically reviewing the study sample

Critical review of IV

Critical Review of DV

Are the instruments/tools chosen to measure the outcome valid and reliable?

if they are valid they are also reliable But if they are reliable it may not be valid

Final Exam Study Guide Flashcards

(191 cards)