Reading...
Play button
Play button
Use LEFT and RIGHT arrow keys to navigate between flashcards;
Use UP and DOWN arrow keys to flip the card;
H to show hint;
A reads text to speech;
323 Cards in this Set
- Front
- Back
|
Ways we care about our patients:
|
Establish rapport
Not being judgmental Positive attitude Insure privacy Show concern Other psychological variables Providing the best clinical intervention possible. |
|
|
How do we provide the best clinical intervention possible?
|
Don't have to accept standard procedures if you have doubt or concerns
To truly provide the best clinical intervention, you have to be knowledgeable about the intervention you're using. Requires an appreciation of the measurements used in evaluating patient outcomes, how they're statistically analyzed, and how to interpret the stats. |
|
|
Assessments (tests):
|
provide the data from the measurements. Use them to evaluate changes in your patients for some variable such as cognition, strength, mobility, balance, etc.
|
|
|
What does an assessment provide?
|
Data (measurements) which is statistically evaluated to make a value judgment or the initial variable you're concerned with.
|
|
|
Measurement:
|
the act or process fo quantifying some variable.
|
|
|
How do we quantify a variable?
|
You have to collect data which means you have to measure something and decide what to measure.
|
|
|
Big Picture: What you want to evaluate -> ____ -> ____ -> ____ -> evaluation and decide what to do
|
assessments (tests); measurements (data); statistical analysis
|
|
|
For all assessments, are you measuring variables or degrees of difference?
|
Degrees of difference. Data are then statistically analyzed and used to make decisions about the initial variable.
|
|
|
Critical thinking:
|
The art of analyzing and evaluating thinking w/ a view to improve it.
|
|
|
Advantages of being a critical thinker:
|
You can raise vital questions and problems, formulating them clearly and precisely
You can gather and assess relevant info, using abstract idease to interpret it effectively. You can come to well-reasoned conclusions and solutions, testing them against relevant criteria and standards. You can think open-mindedly w/in alternative systems of thought, recognizing and assessing, as need be, their assumption, implication, and practical consequence. You can communicate effectively w/ others in figuring out solutions to complex problems. |
|
|
Characteristics of critical thinking:
|
Self-directed
Self-disciplined- teach yourself Self-monitored- have to control it. Self-corrected Involves effective communication. Involves problem solving abilities- follows a scientific method. Involves a commitment to overcome our negative egocentrism and sociocentrism. |
|
|
Egocentrism:
|
the tendency to perceive, understand, and interpret the world in terms of the self.
|
|
|
Sociocentrism:
|
The tendency to perceive, understand, and interpret the world in terms of your society, culture, or profession.
|
|
|
Examples of egocentrism and sociocentrism:
|
We don't naturally consider the rights and needs of others.
We don't naturally appreciate the point of view of others. We don't naturally appreciate the limitation sof our own points of view. We have to be trained to be aware of our own egocentrism. We don't naturally recognize our own egocentric assumptions when reasoning. We don't naturally recognize our self-serving perspective. |
|
|
Innate egocentrism:
|
It's true b/c I believe it.
I assume what I believe is true even though I haven't questioned the reasons for my belief. |
|
|
Innate sociocentrism:
|
It's true b/c we believe it.
I assume the dominant beliefs w/in the groups that I belong are true even though I haven't questioned it. |
|
|
Innate self-validation:
|
It's true b/c I want to believe it
I believe what feels good, supports my other beliefs, what doesn't require me to change my thinking in any significant way, what doesn't require me to admit that I've been wrong. It's true b/c I've always believed it. Gives them support, balance, and security. |
|
|
Innate selfishness:
|
It's true b/c it's in my selfish interest to believe it.
Hold fast to beliefs that justify my getting more power, money, or personal advantage even thoguh these beliefs aren't based on sound reasoning or evidence. |
|
|
"healing by coincidence"
|
Healing due to variability (natural history) of the disease itself.
|
|
|
Scientific method:
|
systematic inquiry as to the nature of the universe which follows a pattern.
|
|
|
Steps of Scientific measure:
|
1. empirical data is generated
2. Generation of a hypothesis. 3. Experiments (treatments) are conducted and data (measurements) are generated. 4. The data is statistically analyzed. 5. Based on the interpretation of the statistics, the hypothesis is rejected or accepted. 6. Theories may be generated. |
|
|
How is empiracal data generated?
|
Look at something you want to analyze.
Make observations, take measurements, based on experience. Must realize empirical should be objected, not subjective. Needs to be facts, not interpreted. |
|
|
2 types of hypotheses:
|
Null and alternate hypothesis
|
|
|
Null hypothesis:
|
no difference exists between two explanations
|
|
|
Alternate hypothesis:
|
a difference does exist between the two explanations.
|
|
|
When are theories generated?
|
When a hypothesis is repeated multiple times by different people and is shown to be true. Fundamental theories will be found untrue over time.
|
|
|
Two paths of reasoning followed by the scientific method:
|
Inductive and deductive reasoning
|
|
|
Inductive reasoning:
|
Specific thought to general thought. Reasoning from individual cases or specific facts and coming to a general conclusion.
|
|
|
Deductive reasoning:
|
General thought to specific thought.
Reasoning from a general idea, principle, or thought to something specific. More common in science and clinical research than inductive reasoning. |
|
|
Models:
|
A series of steps or events which explains a process.
|
|
|
What does it mean when you say models are paradigms waiting to change and thus are outdated once they are born?
|
Once you create a model there are variables that will eventually change and will influence a model.
|
|
|
Paradigm:
|
how we think about things.
|
|
|
2 Disablement models:
|
International Classification of Impairments, Disabilities, and Handicaps (ICIDH)
Nagi Scheme |
|
|
ICIDH components
|
Disease
Impairment Disability Handicap |
|
|
Nagi components
|
Active pathology
Impairment Functional limitation Disability |
|
|
Research:
|
An objective, systematic investigation.
A scientific process of inquiry and/or experimentation that involves purposeful systematic and rigorous collection of data. Analysis and interpretation of the data are then made in order to gain new knowledge or add to existing knowledge. |
|
|
Research paradigm:
|
Descriptive data you cannot infer to the future.
|
|
|
How do you come up with rehab research ideas?
|
Clinical observations
Asking "why do we do that?" Ask yourself if this makes sense? Associate what you're observing w/ what you know. Look for areas w/ limited research. |
|
|
Types of research papers you'll be reading as a clinician:
|
Quantitative and qualitative research.
Basic/pure/bench research Clinical research Applied research Methodological research Descriptive research Epidemiological research |
|
|
Quantitative:
|
considered a higher level of research than qualitative since the results can be generalized to the general population and qualitative cannot.
|
|
|
Basic/pure/bench research
|
Quantitative
Deal w/ establishing new knowledge in the development or refinement of theory. |
|
|
Clinical research
|
Quantitative
Articles that involve human subjects that receive different treatments. Type of applied research directed toward solving clinical problems by conducting clinical trials of new programs. Ex: use of support hose to prevent DVT |
|
|
Applied research
|
Quantitative
Designed to answer practical problems. Ex: development of MRI machines |
|
|
Methodological research
|
Quantitative
To develop or design new ways to measure changes between variables. All tests and assessments are methodological types projects. |
|
|
Methodological research is based on 4 things:
|
validity, reliability, sensitivity, and predictability
|
|
|
Descriptive research
|
Qualitative- only meaningful for the population that it's describing, can't be generalized to other populations.
Designed to describe systematically a situation, condition, observation, or area of interest cannot use any of the information and put it towards something else. Ex: characteristics of entering DPT class can't be generalized to nursing. |
|
|
Epidemiological research
|
Qualitative in nature, but can be quantitative.
Designed to study the incidence, distribution, cause of disease, or impairment. |
|
|
Research methodology:
|
determines how you set up your experiment/research to evaluate the null hypothesis.
Involves deciding you're going to measure. |
|
|
What is quality of research based on?
|
How many of the 3 concepts of research methodology you incorporate.
|
|
|
3 Important concepts that research methodology revolves around:
|
1. Manipulation
2. Control 3. Randomization |
|
|
Manipulation:
|
The researcher is changing (manipulating) one or more variables in connecting the subject or condition.
|
|
|
Variable:
|
anything that can vary or change about the condition or subject being tested.
|
|
|
2 types of variables:
|
Dependent and independent
|
|
|
Independent variable
|
The variable which is manipulated which can be the experimental intervention or treatment variable.
Ex: hot/cold; drug |
|
|
Dependent variable
|
The data (measurement) outcome, condition, or appearance varaible.
Ex: swelling of ankle; confusion. |
|
|
Control:
|
Refers to the ability of the researcher to control or eliminate interfering and irrelevant influences from the study. Want dependent variables to be as pure as possible.
Need to be able to say results are due to manipulation of independent variables of the study not chance interference of outside variables. |
|
|
If there isn't a control group, what can be used instead?
|
Each subject can be their own control
Ex: fMRI project, there were periods where subject didn't squeeze ball that could be compared to activity. |
|
|
Randomization
|
A process designed to reduce the risk of systematic bias from creeping into the study. Ensures subjects are representative of the group from which they are chosen and that groups are similar.
|
|
|
Internal validity:
|
The chance we're changing and measuring what we think we're changing and measuring. Maximized through randomization.
|
|
|
External validity:
|
The chance results found in subjects can be applied to groups outside of the groups we're studying.
|
|
|
Research protocols fall into what 3 categories?
|
True experimental design
Quasi-experimental design Non-experimental design (descriptive/qualitative research) |
|
|
True experimental design must have:
|
Manipulations (variables)
Control group Randomization |
|
|
True experimental design is what type of research?
|
cause-effect research
|
|
|
Double-blind research
|
type of true experimental design. Both subjects and researcher are blind to independent variable.
|
|
|
Quasi-experimental design:
|
Require manipulation of the independent variable, but don't have controls or randomization. Since no randomization or controls, study is open to outside influence and isn't as accurate.
Includes case studies. Considered pre-experimental design to see if null needs to be further investigated. |
|
|
Ex of quasi-experimental design:
|
Pretest, treatment, post test design where pretest measure is the control value.
|
|
|
Non-experimental designs (descriptive/qualitative research):
|
No manipulation of independent variables, no controls, and no randomization.
Basically describing characteristics and events connected w/ sample populations, individual, etc. Describing fixed data. |
|
|
Benefits of non-experimental designs:
|
Generates questions for further research.
Good correlational studies. |
|
|
Data collection (measurements):
|
The numeric value assigned to an object, event, interaction, observation, or person according to rules.
Become the operational criteria of the assessment, which tells you how you're going to measure something. |
|
|
Objective of measurements:
|
to standardize criteria or rules so the dependent variables are measured the same.
|
|
|
Categories of measurements based on purpose:
|
Fundamental measurement
Derived measurement Change measurement |
|
|
Fundamental measurement:
|
Obtained initially w/o the need for derivation (no math calculations). Straight forward, what ever the measurement is is what it is.
|
|
|
Derived measurement:
|
Measurements of a variable (dependent) that are obtained as a result of a math operation applied to the existing fundamental measurement.
|
|
|
Change measurement:
|
Mathematical difference between 2 of the same kinds of measurements taken on the same person at 2 points in time.
Ex: pre and post treatment data. |
|
|
Purposes of measurements in clinical research:
|
Evaluative purpose
Predictive purpose Discriminative purpose |
|
|
Evaluative purpose:
|
Purpose is to evaluate the effect of an intervention over time; want to know if patients are getting better. Sometimes referred to as outcome measures.
|
|
|
Predictive purpose:
|
Using a measurement to discriminate something about future events or conditions such as predicting the prognosis.
|
|
|
Discriminative purpose:
|
Using a measurement to discriminate some function, variable, or activity among subjects or groups.
|
|
|
What different levels do assessments/measurements occur at?
|
Active pathology (cellular) level
Activity or systems level Functional level Disability/handicap level |
|
|
Active pathology (cellular) level
|
Tissue level; all kinds of chemical and blood tests; muscle enzyme tests
|
|
|
Activity or systems level
|
The impairment level (what's not working), walking or gait, etc.
|
|
|
Functional level
|
Assessment that evaluates patients ability to run, walk, jump, reach, etc.
|
|
|
Disability and handicap level:
|
Limits in performance of socially defined roles and tasks within a social, cultural, and physical environment.
|
|
|
Types of data measurements
|
Qualitative (alphanumeric) data
Quantitative data (numeric/quantity) |
|
|
Qualitative data:
|
Data which is composed of letters or characters which may be expressed as digits.
Sometimes called character or categorical data. If a digit is assigned, it's numerical value is meaningless. Nominal numbers. Use descriptive stats to describe qualitative data (means, modes, etc) |
|
|
Nominal numbers:
|
Numbers on a jersey, zip code, room number, etc.
|
|
|
Quantitative data:
|
May be expressed as decimals, positive, or negative.
In order to generate it, there must be measurements. Discrete or cardinal variables. Continuous variables. |
|
|
Ex of quantitative data:
|
Goniometry, BP, pulse rate
|
|
|
Measurement:
|
The act or process of quantifying events, people, reactions, characteristics, etc.
It's important the measurement is standardize or it won't lend itself to statistical analysis. Numbers generated reflect how much of an attribute (dependent variable) is present or absent. |
|
|
Discrete or cardinal variables:
|
If the variables can only be described in whole numbers.
Used in quantitative data. Can also be used in qualitative data if used w/o value. Can be referred to as count data. |
|
|
Continuous variables:
|
Variables which can take on any value along a continuum w/in a definite range.
Ex: gait speed in feet/minute. |
|
|
What types of data can scales be used for?
|
Quantitative and qualitative
|
|
|
Name the scales in order of importance in terms of research validity.
|
Nominal scale
Ordinal scale Interval scale Ratio scale |
|
|
Which scale is used for qualitative data?
|
nominal
|
|
|
Nominal scale:
|
Lowest level of refinement and least informative.
Measures discrete variables which are associated w/ qualitative data. Useful for sorting items and establishing mutually exclusive groups or categories. No object/person can be assigned to more than one group. No group can be ordered or ranked above another. Descriptive statistics (frequencies, proportions, ratios, percentages, etc) Non-parametric data |
|
|
Ordinal scale:
|
It's a ranking scale, implying a greater/lesser degree of something.
No equality of difference between categories. Uses descriptive statistics. Symbols assigned to categories are arbitrary, but have to preserve the ranking system. Data is discreet and non-parametric. |
|
|
Interval scale:
|
Data is ranked in a logical sequence and the intervals between sequences of numbers are considered equal and represent actual values.
Since data is on a continuum, relative difference and equivalence w/in scale can be determined. No absolute zero on scale. If researcher assigned zero, it's arbitrary. Used to measure quantitive data. Used w/ parametric data. |
|
|
Ratio scale:
|
Highest level of measurement
Continuum of values. Basically the same as interval scale, except it has an absolute zero point which has empirical meaning. There are absolute values |
|
|
Statistics:
|
to extract the maximum info about a set of data. This can be both qualitative and quantitative.
|
|
|
Population:
|
The total number of individuals, measurements, or data from which measurements will be collected or generalized about.
|
|
|
Sample of the population:
|
portion or subset of the population
|
|
|
Test statistic:
|
when you're using a statistic that's specific for samples
Ex: FIM score to asses disability. Score is a test statistic |
|
|
3 levels of data analysis:
|
Descriptive analysis
Correlative or trend analysis Comparative analysis |
|
|
Descriptive analysis
|
Lowest level of data analysis
Describe something w/ numbers: mean, mode, etc. of height, weight, etc. Qualitative research |
|
|
Correlative or trend analysis:
|
Describes relationship of changes of one variable w/ changes of another variable.
Middle level. Use things called correlation coefficients. Can extrapolate results to population. Quantitative research |
|
|
Comparative analysis
|
Determines whether 2 or more groups of data are different or not.
Cause and effect relationship Highest level. |
|
|
With samples, you have 2 categories of tests:
|
Parametric and non-parametric statistical tests
|
|
|
Parametric statistical tests:
|
test which are run when the data comes from a normal distribution (bell shaped curve).
|
|
|
Bell shaped curve:
|
has data that clusters around the mean.
68.2% of data falls w/in one SD above or below. Mode and median are also values around the mean. Can be used w/ larger populations. |
|
|
Non parametric statistical test:
|
Tests which are run when data don't come from a normal distribution and you can't do a parametric test.
|
|
|
When a normal distribution cannot be assumed, examples of non parametric tests include:
|
Chi squared tests used to compare observed frequencies to expected or theoretical outcomes.
Spearmen rank correlation coefficient. |
|
|
Parametric and non-parametric data can involve what types of statistical tests?
|
Descriptive and inferential statistical tests.
|
|
|
Descriptive statistics:
|
Those which describe, organize, and summarize data.
Cannot compare things or extrapolate data to the general population. Lower level of research. Not as valid as inferential b/c can't infer to population. |
|
|
Descriptive stats include:
|
Frequency
Percentages Percentiles Prevalence Incidence Descriptions of central tendency Description of relative position |
|
|
Frequency:
|
the number of occurrences of a repeating activity based on a unit of time.
|
|
|
Percentages:
|
A way of expressing a number as a fraction of 100
|
|
|
Percentiles
|
Values such that a specified percent of data falls above or below a value.
|
|
|
Prevalence:
|
the total number of cases in the population or sample at a given time or point in time.
Usually expressed as a percentage. Used as an estimate of how common a condition is w/in a specific population at a particular point in time. |
|
|
Incidence:
|
Measurement of the number of new individuals who develop a disease or condition w/in a particular period of time.
Normally expressed as a percent. |
|
|
Descriptions of central tendency:
|
mean, median, mode
|
|
|
Mean:
|
arithmetic average
|
|
|
Median:
|
value which separates a sample from the upper and lower half
|
|
|
Mode:
|
Value which occurs most frequently
|
|
|
Description of relative position:
|
range, SD, and SEM
|
|
|
Range
|
Interval between minimum and maximum values
|
|
|
Standard deviation:
|
Measure of the variability of a population, sample, or probability distribution.
Want them to be low values b/c we're measuring variations. Low= closer to mean. Use same units as the data you're measuring. |
|
|
SEM
|
quantifies the certainty w/ which the mean computed from a random sample estimates the true mean of the population from which the sample was drawn.
Measures variability. More accurate than SD. |
|
|
Inferential statistical tests:
|
Test which use data from samples drawn from a population to make inferences about the total population.
|
|
|
Types of inferential statistical tests:
|
Student t-test
Analysis of Variance (AOV/ANOVA) |
|
|
Student t-test:
|
Parametric statistical test
|
|
|
2 types of student t-test:
|
Unpaired and paired
|
|
|
Unpaired student t-test:
|
Used to test whether the mean drawn from a normal population differs from hypothesized value.
|
|
|
Paired student t-test:
|
used to test whether the means of 2 groups are different, when the samples were drawn in paires (pre and post) or are related.
|
|
|
Analysis of variance
|
An extensive class of related statistical models and their associated procedures in which the observed variance (SD/SEM) is separated into categories due to different independent variables.
|
|
|
One-way analysis of variance
|
statistical tests which normally involve 3 or more independent variables and only 1 dependent.
Ex: TKA patients w/ 3 intervention options. |
|
|
Correlation coefficients:
|
an index of the degree of association between 2 variables or the extent to which the order of individuals on 1 variable is similar to the order of individuals on the 2nd variable.
Strong correlations of data don't necessarily prove cause and effect. Want r as close to one as possible. |
|
|
Linear regression analysis:
|
establishes a math relationship between 2 or more variables.
|
|
|
Types of correlation coefficients:
|
Pearson's correlation coefficient
Interclass correlation coefficient Spearmen rank correlation coefficient Cronbach's alpha Cohen's kappa coefficient |
|
|
Pearson's correlation coefficient:
|
Quantifies the strength of association between 2 variables that are normally distributed.
Parametric Usually shown by "r" in papers Normally used w/ true experimental design Type of ICC |
|
|
Interclass correlation coefficient (ICC)
|
Demonstrates the consistency of measurements when one or more raters takes the measurements.
Type of inferential statistical test used with parametric data. |
|
|
Spearmen rank correlation coefficient:
|
Used to quantify the strength of association between 2 variables measured on an ordinal scale.
Want to see if variables are related in a correlation in a positive/negative manner. Non-parametric test. Marked by Greek letter rho. |
|
|
Cronbach's alpha:
|
Frequently used as a measure of the internal consistency reliability of an assessment.
How close the dependent variable is to being close each time you measure. How accurate. |
|
|
Cohen's kappa coefficient
|
A measure of inter-rater agreement for qualitative items.
Only used for qualitative. Can only involve 2 raters. Want high positive or negative numbers. |
|
|
Validity:
|
Refers to the degree to which a test, intervention, or instrument measures what it's supposed to be measuring.
Not an all or none property, but a degree or spectrum. It's a property of the measurement, not the assessment/measuring device. |
|
|
External validity:
|
implies that we can generalize the results of an assessment to similar populations.
|
|
|
Internal validity:
|
Concerned w/ correctly concluding that an independent variable is, in fact, responsible for variation in the dependent variable.
Cause and effect: validity will be high. Need to use controls to have good internal validity and should randomize and manipulate. |
|
|
What might cause you to loose internal validity?
|
History
Multiple testing causes learning curve on how they perform. Selection of subjects and randomization. Instrumentation not calibrated right. |
|
|
Construct validity:
|
Based on the knowledge and intellectual underpinnings, which are considered the construct upon which the test and measurements are developed.
|
|
|
Content validity:
|
Related to the extent to which a measurement reflects the specific intended domain of content.
Type of non-statistical validity. |
|
|
Criterion-based validity:
|
Involves comparing the measurements being examined w/ another measurement or series of measurement or procedures which have been demonstrated to be valid.
|
|
|
3 types of criterion-based validity:
|
Concurrent validity
Predictive validity Prescriptive validity |
|
|
Concurrent validity
|
When an inferred interpretation is justified by comparing a measurement w/ supporting evidence that was obtained at approximately the same time as the measurement being evaluated. More precise b/c of time frame.
|
|
|
Predictive validity:
|
concerned w/ using criterion to make predictions which are true.A lot of screening tsets have this, if we don't screen for certain things we can almost predict something will go wrong if left untreated.
|
|
|
Prescriptive validity:
|
Concerned w/ using the inferred interpretation of criterion from a test to prescribe a treatment.
|
|
|
Face validity:
|
Concerned w/ how a measure or assessment appears.
More of a non-statistical type of validity. Does data seem reasonable? |
|
|
Convergent validity:
|
Refers to the degree to which a measure is correlated w/ other measures that it's theoretically predicted to correlate with.
|
|
|
Reliability
|
The degree to which measurements of a test remain consistent over repeated test of the same subject under identical condition.
The degree to which measurements are error free and the degree to which repeated measurements will agree. |
|
|
Types of reliability:
|
Inter-tester reliability
Intra-tester reliability Test-retest reliability |
|
|
Inter-tester reliability:
|
consistency of measurements when more than one person takes the measurements.
Measuring reliability of testers, not the test. |
|
|
Intra-tester reliability:
|
Consistency or equivalence when one person repeated measurements over a period of time.
|
|
|
Test-retest reliability:
|
Consistency of repeated measurements in time.
Indicates stability of reliability over time. Measuring the tool/assessment. |
|
|
True positives:
|
Sick subjects correctly identified as having the disease.
|
|
|
False positives:
|
Healthy subjects wrongly identified as having the disease.
|
|
|
True negatives:
|
Healthy subjects correctly identified as not having the disease.
|
|
|
False negatives:
|
Sick subjects incorrectly identified as not having the disease.
|
|
|
Sensitivity:
|
A value which indicated the proportion of actual positives which are correctly identified as being positive.
Identified those who had a disease who actually had it. Same as TRUE POSITIVE. |
|
|
Sensitivity formula:
|
Number of true positives/(number of true positives+ number of false negatives)
|
|
|
Specificity
|
A value which indicates the proportion of negatives which are correctly identified as being negative.
Correctly identified those subjects w/o the disease who don't have it. Same as TRUE NEGATIVE |
|
|
Specificity formula:
|
Number of true negatives/(number of true negatives+ number of false positives)
|
|
|
APTA Standards of measurement:
|
A well constructed test is only valid when used for the purposes for which it has been developed.
|
|
|
Without reliability, there's no ___.
|
Validity
|
|
|
Psychometric properties (questions they answer):
|
How well do items on a test define a construct of interest?
Does the item set reflect a useful range of performance? How consistent are the test scores? How well does the score on the test measure a construct of interest? How well does the score on the test measure the performance of interest? What's the clinical significance of the test score? What are the consequences of making specific decisions based on the score? |
|
|
Psychometric attributes of standardized tests:
|
Reliability
Validity Diagnostic efficiency |
|
|
Intrasubject variability:
|
variations w/in the individual being assessede
|
|
|
Validity:
|
Test score measures what we intend it to measure.
Construct Responsiveness Reliability change index |
|
|
Responsiveness:
|
Characteristic of validity. The score's ability to indicate the presence of change when real change occurs. Has nothing to do w/ the diagnosis, but change in test scores.
|
|
|
Diagnostic efficiency:
|
Sensitivity
Specificity Positive and negative predictive values |
|
|
Reliability:
|
The extent that a test measure is accurate, consistent and free from error.
Consistency of the measurement and people using it. SEM |
|
|
Intra and interrater:
|
variations within or among observers
|
|
|
SEM:
|
Index of the dispersion of an obtained score about a true score correlation coefficient > .90 for diagnostic decisions and >.80 for screening decisions.
|
|
|
Validity (Kalobe lecture)
|
Consistency w/ theoretically derived hypotheses concerning the construct that's being measured.
The extent to which test scores measure what is intended to measure. Extent to which the measure produces the same results as the criterion measure (gold standard). Meaningfulness of the interpretations made based on test scores. Ability of the scale to correctly identify those with a target condition. Proportion of individuals classified as abnormal by the criterion measure, who were classified as such by the scale. |
|
|
Diagnostic validity:
|
sensitivity
|
|
|
Types of validity:
|
Face
Content Construct Concurrent Predictive Ecological |
|
|
Ecological validity:
|
Correlates to what actually occurs in the environment the person normally is in.
|
|
|
Cut-off score:
|
Subjective points used to base decisions about whether or not a person:
Has a condition Is eligible for a specified intervention Needs to be referred for further testing. |
|
|
The decision about the cut-off score is a balancing act...
|
How much error is one willing to live with?
|
|
|
Standard scores:
|
Z-score
Scaled score Percentile Age equivalent score |
|
|
Percentile rank:
|
A cut-off point.
Position of individual's score relative to the normative sample. Getting a percentile based on your ranking against someone else. |
|
|
Age equivalent score:
|
The age derived from chronological age of children in the normative sample whose average raw score is the same as the raw score achieved by the individual.
Developmental score/chronological age. |
|
|
Purpose of cut-off scores:
|
Diagnosis/discrimination
Predict outcome Evaluate change or progress |
|
|
How good is a cut-off score? 3 types:
|
Arbitrary- tendinitis, pain
Norm-referenced- mathematical cutoff score. How far are you from mean? (SD, percentile) Criterion- referenced- 75%, A+ |
|
|
What to consider in setting a cut-off score:
|
Level of competence required to perform the task (criterion).
Not set below what would correspond to an acceptable level of performance of task. Sample size and type. |
|
|
Standard normal distribution (percentages on bell curve):
|
1 SD: 34.1% on each side (68% total 1SD above and below mean).
2 SD: 13.6% on each side 3 SD: 2.1% on each side |
|
|
Positive predictive values:
|
A proportion of individuals identified by the cutoff score as being abnormal who are classified as having a target condition by a criterion measure.
|
|
|
Negative predictive value:
|
A proportion of individuals identified by the cutoff as being normal who are classified as not having a target condition by a criterion measure.
|
|
|
Sources of instability of sensitivity and specificity:
|
Verification bias
Small samples- # w/ a target condition= large confidence intervals. Very susceptible to outliers. Errors in criterion measure (diagnostic test). Construct irrelevant variance- unrelated to test construct. |
|
|
Verification bias:
|
Differential loss of either positives or negatives (100 vs. 70 for criterion)
|
|
|
What every clinician needs to know about the population of interest:
|
Setting data, census, or literature
|
|
|
The higher the prevalence, the ___ the PPV and vice versa.
|
higher
|
|
|
If have a high prevalence of what you're measuring, the cutoff point will be ___ and favor what the sample is based on.
|
Higher
|
|
|
Likelihood ratio:
|
Incorporates sensitivity and specificity. Much more accurate, but more tedious to calculate.
Provides a direct estimate of how much a positive or negative test result will change the likelihood of having the condition or disease. |
|
|
Positive result (LR+):
|
How much the likelihood of the condition increases when a test is positive.
|
|
|
Negative result (LR+):
|
How much the likelihood of the condition decreases when a test is negative.
|
|
|
Positive likelihood ratio formula:
|
LR+ = sensitivity/ (1-specificity)
|
|
|
Negative likelihood ratio formula:
|
LR- = (1-sensitivity)/specificity
|
|
|
Responsiveness and cutoff score:
|
The power of a scale to detect a difference when one is present.
|
|
|
Magnitude of change that is clinically significant measured w/:
|
Minimal detectable change (MDC)
Minimal Important Difference (MID) or Minimal clinically significant difference (MCSD) |
|
|
Minimal detectable change:
|
Minimal change that could be attributed to intervention vs. error
Related to test reliability: SEM |
|
|
Minimal important difference:
|
Clinically relevant change; magnitude of change that's meaningful or change in function beyond natural progression.
|
|
|
Clinical significant change
|
A change that's recognizable to peers and recognizable to others.
A proportion of persons who show improvement (or deterioration). A proportion of elimination of the presenting problem. Minimally clinically significant change- must relate to range of change, not dichotomy. |
|
|
ROC:
|
measure of the probability that the perceived abnormality will allow correct identification. Can get the same info from sensitivity, specificity, etc.
More complicated. |
|
|
Item response theory and cutoff score:
|
Rasch model0 cultural validity. Tells the same thing in a population other than which this measure was developed for.
|
|
|
Assessment purposes:
|
Diagnostic
Evaluate change Predict |
|
|
Interpretation:
|
A process of assigning meaning to the collected information and determining significance or implications of the findings.
Basis for clinical decision making. |
|
|
Steps of interpretation:
|
Purpose of assessment
Type of information collected Questions to be answered |
|
|
For a normal standard distribution, what percent will score within 1 SD above or below mean?
|
68%
|
|
|
What percent will score within 2SD above or below mean?
|
95% (95% confidence interval?)
|
|
|
Why is knowing the mean and SD important?
|
Knowing the mean and standard deviation makes it possible to interpret raw scores and compare different individuals' performances
performance on one test with his or her performance on another test. W/o standardized scores, it's difficult to make comparisons |
|
|
Interpretation of scores can be done using:
|
Norms (standard scores)
Percentile rank Age equivalent Criterion scores |
|
|
Standard deviation:
|
dispersion/variance of scores
|
|
|
z score:
|
Where a value stands within the dispersion. How many SDs a person is above or below the mean.
|
|
|
If the mean is 100 w/ a SD of 15, then 2/3 of responses lie between what values?
|
85 and 115
|
|
|
A z-score of -1.4 means what?
|
1.4 SD below the mean
|
|
|
Z score of 1 in the above example=
|
1 SD of 15
|
|
|
SEM=
|
sV(1-r)
V means square root. |
|
|
In the SEM =, s=?
|
SD for the test
|
|
|
In the SEM=, r=
|
reliability coefficient for test.
|
|
|
The larger the SEM, the ___ reliable the test.
|
less
|
|
|
Sensitivity
|
Ability of the scale to identify correctly those w/ or w/o the target condition.
|
|
|
Specificity
|
Proportion of individuals classified as abnormal (or normal) by the criterion measure, who were classified as such by the scale
|
|
|
Sensitivity and specificity answer the following question:
|
Of the individuals that the criterion measure identified as having a conition, how many were identified as such or normal by the scale one wants to use?
Help guide decision on whether or not to choose a tool. |
|
|
Positive and negative predictive value
|
Proportion of individuals identified by the cutoff as being abnormal/normal who are classified as having a target condition by criterion measure
|
|
|
Pos and Neg predictive values answer what question?
|
Of individuals who the tool one is using IDed as having a condition, how many did indeed have the condition?
|
|
|
PPV and NPV are concerned with what?
|
Interpretation of test results
|
|
|
Diagnostic validity includes:
|
Sensitivity and specificity and NPV/PPV
|
|
|
Mindy was referred to you for a consultation. One of the tests you have administered to her has a reliability coefficient of .96 and SD= 15. What's the SEM?
|
SEM= 15V(1-.96)= 15V(.04)= 15x.2= 3
|
|
|
The larger the SEM the ___ reliable the test.
|
Less
|
|
|
Chances are 68/100 that her true score falls between:
68% CI |
97 and 103?
|
|
|
Chances are 95/100 (95% CI) that her true score falls between:
|
94-106?
|
|
|
95% CI formula:
|
Test score + 2(SEM)?
|
|
|
SD formula:
|
SD= Vsum (x-m)squared/(n-1)
x= scores m= mean n= number of values |
|
|
Sensitivity= .59, specificity= .86, PPV= .39, NPV= .77.
T/F: 59% of individuals who were found to have hip fracture were identified as so by the test. |
True
|
|
|
Sensitivity= .59, specificity= .86, PPV= .39, NPV= .77.
T/F: 39% of individuals who were found to have hip fracture were identified as such by test. |
False
|
|
|
Sensitivity= .59, specificity= .86, PPV= .39, NPV= .77.
T/F: 86% of individuals who were identified by the test as being normal were later classified as such. |
False- 77%
|
|
|
Sensitivity= .59, specificity= .86, PPV= .39, NPV= .77.
T/F: Of the individuals confirmed to be normal, 77% were correctly identified as such by the test. |
False- 86%
|
|
|
Sensitivity= .59, specificity= .86, PPV= .39, NPV= .77.
T/F: It would be better to retest the elderly periodically b/c the low PPV indicate a high rate of false positives. |
False. It's true that you have too many false positives, so you should NOT retest them.
|
|
|
Sensitivity= .59, specificity= .86, PPV= .39, NPV= .77.
T/F: It would be better to retest the elderly periodically b/c the low PPV indicates a high rate of false negatives. |
True
|
|
|
Sensitivity= .59, specificity= .86, PPV= .39, NPV= .77.
T/F: Using this test may result in over-referral of the elderly to PT. |
False If PPV is low, the people that you say are likely to fall aren't likely to fall. You shouldn't be testing them always or periodically testing them, they are not the issue, your tool is.
|
|
|
If sensitivity is .25, is there a high incidence of false negatives or false positives?
|
False negatives b/c the tool won't pick up everyone it should
|
|
|
High sensitivity-
|
increased incidence of false positives
|
|
|
Low sensitivity-
|
Increased incidence of false negatives
|
|
|
High specificity-
|
Increased incidence of false negatives
|
|
|
Low specificity
|
Increased incidence of false positives
|
|
|
Percentile rank:
|
Position of an individual's score relative to the normative sample.
|
|
|
A patient's percentile rank on a measure of vital capacity is 15. How would you interpret this?
|
Scored better than 15% of those test.
|
|
|
Test for hip subluxation has 92% sensitivity, 86% specificity. What's the likelihood ratio for a positive result?
|
LR += .92/(1-.86)= 6.57= 6.6
|
|
|
Likelihood ratio:
|
Incorporates sensitivity and specificity
Provides a direct estimate of how much a positive or negative test result will change the likelihood of having a condition or disease. Need to know prevalence. |
|
|
Positive result (LR +)=
|
How much the likelihood of the condition increase when a test is positive.
sensitivity/1-specificity |
|
|
Negative result (LR-)=
|
How much the likelihood of the condition decrease when a test is negative.
LR-= 1-sensitivity/specificity |
|
|
Test for hip subluxation has 92% sensitivity, 86% specificity. What's the likelihood for a negative result?
|
1-.92/.86=.09
|
|
|
Clinical significance change:
|
A change that's recognizable to peers and others.
Change in daily functional activities. A proportion of persons who show improvement or deterioration. Proportion of elimination of the presenting problem. Must relate to change not dichotomy (no overlap) |
|
|
29 month old child w/ CP whose age equivalent score on the Peabody changed from 18 to 19 months in fine motor subscale and 16-18 months on following 6 months of therapy. SEM= 3 mos. How would you interpret change score?
|
Initial scores 18-19 months. W/ SEM 3, 68% of the time the child's score fell between 15-21.
|
|
|
Minimal important difference:
|
Clinically relevant change; magnitude of change that's meaningful or change in function beyond natural progression.
|
|
|
T-test: t=4.1, df=2.45, p=.017. Is this good or bad?
|
good
|
|
|
Correlation of .47, good or bad?
|
Bad
|
|
|
Item Response Theory (IRT)
|
the psychometric technology that allows equitable scores to be computed across different sets of items.
Mathematical models that describe how test takers interact w/ items Rasch model |
|
|
Rasch model:
|
Unidimensionality
Item difficulty Interval scale Elimination of misfitting items. Calculation of total scores from administered items |
|
|
Computer Adaptive Testing
|
A method of test administration based on the child's ability level.
Also known as Tailored testing Selects questions to maximize test precision. CAT require fewer test items to arrive at accurate scores. Interactive algorithm method. |
|
|
Steps of CAT
|
Select an item from a pool of available items based on estimated child ability.
Child answers correctly or incorrectly. Computer updates the estimate of the child's ability based on the previous response pattern on the items and selects a subsequent item. Repeat steps until termination criteria is met. |
|
|
RPR-Return to Play Readiness Measures:
|
quadriceps muscle strength and endurance, hamstring muscle strength and endurance, biomechanics of landing from a vertical jump, player’s max velocity, agility, cardiovascular endurance, speed ( meters completed in X seconds)
|
|
|
RPR Evaluates:
|
discriminating whether or not soccer player is ready to return to play (decreased risk of re-injury) ; evaluating soccer player’s progress in rehabilitation
|
|
|
RPR: what types of patients will use:
|
female collegiate soccer players, age 18-25, recovering from a knee injury
|
|
|
RPR: clinical settings
|
outpatient orthopedic clinic, college sports medicine clinic
|
|
|
Quick Reach Screen (QRS) measures:
|
Number of dominoes knocked down
|
|
|
QRS evaluates:
|
Reaching accuracy (functional level)
|
|
|
QRS: Types of patients:
|
8 year old children
|
|
|
QRS: Types of settings:
|
primarily school use
|
|
|
SITS-Students in Terrible Sitting measures:
|
degrees of ROM, manual muscle testing scores, scores based on patient reported habits and observation of posture
|
|
|
SITS evaluates:
|
Core stability, flexibility, and posture
|
|
|
SITS types of patients:
|
Healthy students with low back pain
|
|
|
SITS settings:
|
orthopedic outpatient physical therapy
|
|
|
PAKBA (just a mnemonic of our first names...) measures:
|
Functional balance determinants, such as the ability to maintain postural control during both static and dynamic tasks, amount of assistance needed during uneven surface walking, reach displacement, and transfer and gait speed.
|
|
|
PAKBA evaluates:
|
Evaluates an individual's risk of having a fall due to balance dysfunction displayed during the functional balance algorithm.
|
|
|
PAKBA Types of patients:
|
Individuals age 65 or older s/p CVA without any known vestibular pathology.
|
|
|
PAKBA settings:
|
Initially it could be used in the in-patient setting during immediate s/p CVA rehab, but could be continued during out-patient rehabilitation for continued assessment and evaluation of progress.
|
|
|
SAPC: Swing Assessment of Postural Control measures:
|
he SAPC is a test of functional postural control that measures the distance of forward and backward displacement, resulting from external perturbation of a child with CP on a swing.
|
|
|
SAPC evaluates:
|
change in the child’s response to perturbation .
|
|
|
SAPC types of patients:
|
Inclusion criteria: Children with CP; ages 4-12; meeting the GMFCS Levels 1, 2, or 3.
Exclusion criteria: Children that may meet the inclusion criteria, but have uncorrected deficiency in their vision, vestibular or somatosensory systems. |
|
|
SAPC settings:
|
pediatric clinic, outpatient clinic
|
|
|
Readiness to Drive Assessment (RDA) measures:
|
Patient's ability to return to safe driving following right talocrural joint injury with immobilization
|
|
|
RDA evaluates:
|
Force (for pressure applied to brake pedal), Range of Motion, and Sensation
|
|
|
RDA types of patients:
|
Males and females, ages 16-40, with a right talocrural injury and immobilization and resulting decreased function
|
|
|
RDA settings:
|
outpatient clinic
|
|
|
Diabetic Foot Ulcer Test (DFUT) measures:
|
plantar sensation, ankle joint mobility, ankle brachial index, Achilles tendon reflexes, and plantar pressure.
|
|
|
DFUT evaluates
|
the risk of developing a foot ulcer in patients with diabetes type I or type II who are receiving physical therapy services.
|
|
|
DFUT types of patients
|
Patients receiving physical therapy treatment who are diagnosed with type I or type II diabetes
|
|
|
DFUT settings
|
Inpatient, outpatient, skilled nursing facility, nursing homes, and school settings
* primarily inpatient |
|
|
NARA: Non-contact ACL Risk Assessment measures:
|
Patient History
-Gender -Family history -Previous knee injuries -Sports participation Dynamic Posture through assessing landing mechanics Frontal Plane -Knee valgus -Excess pronation -Lateral bending of the trunk Sagittal Plane -Excess knee flexion -Landing flat footed or on hind foot -Shoulders moving forward past feet Static Posture -Subtalar joint pronation -Knee Extension -Knee Valgus -Q-angle -Femoral Antetorsion -Tibial Torsion -Muscular Balance |
|
|
NARA evaluates
|
Predicts a patient’s risk level for a non-contact ACL injury and can be used to evaluate the change in a patient’s risk before and after implementing a training program
|
|
|
NARA types of patients:
|
Teenagers and young adults age 15-30
|
|
|
NARA settings
|
Orthopedic outpatients clinics, athletic facilities, and gyms (basically any environment that has a plinth/table and a flat, dry surface to set up 12 inch stool)
|
|
|
Return to Vault Gymnastics Ankle Assessment (RVGAA) measures:
|
speed in seconds, degrees of rotation, quality of landing, strength of the ankle to withstand high forces, endurance, range of motion, and proprioception.
|
|
|
RVGAA evaluates:
|
To evaluate the level of function of competitive gymnasts who have sustained ankle injuries and to predict whether these gymnasts possess the necessary range of motion, proprioception, strength, power, and speed to successfully perform their skills and return to pre-injury training status on the vault
|
|
|
RVGAA types of patients
|
Competitive (Level 6-10) female gymnasts’ ages 10-22 after sustaining an ankle injury
|
|
|
RVGAA settings:
|
The gymnast’s gym (or just a gymnastics gym) or outpatient orthopedic physical therapy clinic
|
|
|
Functional assessment of total knee arthroplasty patients FATKAP measures:
|
change measured in points in a patient’s perception of functional ability before and after total knee arthroplasty
|
|
|
FATKAP evaluates:
|
patient progress before and after a TKA
|
|
|
FATKAP types of patients:
|
English speaking, Male and Female patients between the ages of 60-80 who have undergone a TKA
|
|
|
FATKAP settings:
|
Hospitals, outpatient surgery centers and therapy clinics
|
|
|
FADAPS- Functional Acute Discharge Assessment for Post Stroke Patients measures:
|
The FADAPS primarily assists therapists, as part of an interdisciplinary team, to suggest discharge locations to patients with stroke as a result of the conclusions from the evaluation. FADAPS measures functional ability to perform bed mobility, chair transfers, and ambulation.
|
|
|
FADAPS evaluates:
|
the patients ability to perform functional tasks and the amount of assisstance needed.
|
|
|
FADAPS patients:
|
Post stroke patients (70 +/- 11 years old)
|
|
|
FADAPS settings:
|
Acute Hospital Setting
|
|
|
Standardization of tests/assessments
|
Over time, some tests become standardized b/c they've been shown to be consistently valid and reliable.
Make value judgements based on what you want to accept. |
|
|
How are some tests standardized/normalized?
|
W/ mathmatical manipulation resulting in data w/ a distribution more congruent w/ a bell shape.
|
|
|
Transformation of measurements=
|
mathematical manipulation
End up w/ charts. |
|
|
Relative standards
|
If you don't want to use standardize tests, you can use the subject as their own control & not some standardized chart.
Only relevant to that individual patient; NOT relevant to other people in the population. |
|
|
Collective relative standards
|
a group of similar relative standards can be statistically analyzed and extrapolated using inferential stats to the general population. This is done a lot of the time.
|
|
|
Why clinicians don't like standardized tests?
|
Standardization on normal subjects.
May not fit easily into paperwork. May take too long to perform. Tool may ask irrelevant info from patients- excessive data. Some rely on home grown assessments. |
|
|
P value
|
What kind of chance you're willing to take of being wrong.
Not a percentage, but a level of confidence. |
|
|
P<.05
|
Accepting you could be wrong 5 times out of 100.
|
|
|
Type I error (alpha/false positive)
|
You reject the null when the null is true.
|
|
|
Type II error (beta error/false negative)
|
You fail to reject the null when the null is false.
|
