NPTE research and Evidence based Practice

Notes from Therapy ED

indepdent variable

Variable that stands alone and isn’t changed by any other variables

EX: age and time

Dependent variable

variable that is being studied and measured; changes when the indepenedent variable is manipulated

EX: height at different ages

Categorical variables

variables categorized by traits that do not have inherent numerical variables. such as Nominal scale and ordinal scale

Nominal scale

classifies variable or scores into two or more mutually exclusive categories based on a common set of characteristics.

EX: sex, race or ethnicity

Ordinal scale

classifies and ranks variables or scores in terms of the degree to which they possess a common characteristics

EX: manual muscles grades as normal, good, fair, poor, trace, or zero

Continuous variables

characteristic with inherent numerical value. include Interval and ratio scales

Interval scale

classifies and ranks variables or scores based on the predetermined equal intervals; with no set zero

EX: temperature, IQ test

Ratio Scale

classifies and ranks variables or scores based on equal intervals and a true zero point

EX: ROM, Scales for weight, force production using a dynamometer

Cross-sectional research

data are collected on an individual or groups of individuals at a single point in time. Re

Retrospective Research

Individuals or groups of individuals are enrolled in the study at a particular timepoint or following the occurrence of a target event and historical data are collected. the study proceeds backwards in time. Data are typically collected from subject interview or review of records

Prospective Research

individuals or groups of individuals are enrolled in the study at a particular timepoint and followed forward in time to determine if a particular outcome occurs

Descriptive Research

collecting data about conditions, attitudes, or characteristics of subjects or groups of subjects

Examples of descriptive research

case studies, developmental, longitudinal, normative, qualitative research

case studies or clincal reporrs

indepth investigation of an individual, group or institution

develpmental research

studies of behaviors that differentiate individuals ar different levels of age, growth and maturation

normative research

investigates standards of behavior, standard values for given characteristics of a sample

EX: gait characteristics

Qualitative research

seeks to understand concepts opinions, or experiences related to social phenomena and complex human behavior.

Correlational Research

attemps to determine the presence and magnitude of relationships or associations between two or more variables.

IF a correlation is closer to +1 what does that indicate

The variables are positively correlated

IF a correlation is closer to -1 what does that indicate

Variables are inversely related

IF the correlation is 0 than that means

the variables are not related

Experimental Research

Attempts to define a cause and effect relationship through group comparisons.

Describe a true experimental design

includes random assignment into experimental group or control group. all other experiences are held similar

Describe a cohort design

quasi experimental design, subjects are identified and followed over time for changes/outcomes following exposure to an intervention; lacks randomization, may or may not have a control group

Describe a within subject design AKA repeated measures

subjects serve as their own controls; randomly assigned to treatment or no treatment blocks

Hypothesis

tentative and testable explanation of the relationship between variables

Null Hypothesis

states that no relationship exists between variables. any relationship found is the result of chance or sampling error

When is the null hypothesis rejected

When a significant difference was observed between groups or treatments

What does a sample represent

The larger group from which they were selected

Describe random selection

all individuals in a population have an equal chance of being chosen for a study

Describe a systematic selection

Individuals are selected from a population list by taking individuals at a specified interval EX: every tenth name

What is an Effect Size

the magnitude of the differences between sample means, allows a statistical test to find a difference when one really does exist

Describe Generalizability

degree to which a study’s findings based on a sample apply to an entire population

What is validity

the degree to which a test, instrument or procedure accurately measures what it is supposed to or intended to measure

Face validity

Poorest form of validity; the assumption of validity based on the appearance of an instrument as a reasonable measure of a variable

internal validity

degree to which the observed differences on the dependent variable are the direct result of manipulation of the independent variable and not some other variable.

external validity

the degree to which the results are generalizable to individuals (general population) or environmental settings outside of the experimental study

Content Validity

degree to which an instrument measures an intended content area. determined by expert judgement. requires both item validity and sampling validity

concurrent validity

the degree to which the score on one test are related to the scores on another criterion test with both tests being given at relatively similar times. Involves comparison to the gold standard.

Predicitve validity

degree to which a test is able to predict future performance

contruct validity

the degree to which a test measures an intended hypothetical abstract concept (non-observable)

Describe Reliability

degree to which an instrument measures a phenomenon accurately, dependable, time after time and without variation

INTERrate reliability

The degree to which two or more indepenet rate can obtain the same rating for a five variable

INTRArater reliability

The degree to which one rater can obtain the same rating for a given variable on multiple measurements trials

Describe sensitivity

Tests ability to correctly identify the proportion of individuals who truly have a disease or condition (TRUE POSITIVE). Help to rule out a condition when the test is negative and the condition is not present

Describe Specificity

Test’s ability to correctly identify the proportion of individuals who do NOT have a disease or condition (TRUE NEGATIVE) Help to rule a condition in when the test is positive and the condition is present

equation for +LHR

Sn / (1-Sp)

Equation for -LHR

(1-Sn) / Sp

describe how to utilize a +LHR

a +LHR close to or greater than 10 indicates that a disease is present while a +LHR closer to 1 means a test does not change the odds

describe how to utilize a -LHR

A -LHR <0.1 indicates that disease is absent while a score closer to 1 means the test has little effect.

define Mean and what it is used for

average; appropriate for interval or ratio data

Define mode and what it is used for

midpoint; 50% of scores are above and below; appropriate for ordinal data

define Mode and what it is used for

most frequently occuring score; appropriate for nominal data

Describe Standard Deviation (SD) and what it is used for

a determination of variability (Difference) of scores from the mean.

Subtract each score from the mean, square each difference, add up all the squares and divide by the number of scores. appropriate with interval or ratio data

Describe a normal distribution

symmetrical bell shaped curve; most scores are near the mean within one standard deviation

Standard error

expected chance variation among thee means, result of sampling error

Type I error

the null hypothesis is rejected by the researcher when it is true

Type II error

the null hypothesis is not rejected by the researcher when it is false.

How do you decrease the occurrence of Type I and Type II error

By increasing the statistical power through increasing sample size, increasing the effect size, increasing the alpha level and decreasing observed variance

Describe the strength of relations with positive and negative correlations

Positive correlations (0 to +1) indicates positive correlation so as X increases so does Y. While a negative correlation ( 0 to -1) indicates an inverse relationship. so as X increases Y decreases.

Good reliability >0.75

Moderate reliability =0.50-0.75

Poor reliability <0.50

If examing for relationships (correlations)

If examining for difference between groups

describe a T test

test of significance used to compare two independent groups created by random assigment and identifiy a difference at a selected probability level

Describe ANOVA

parametric test used to compare three or more independent reatment groups or conditions at a selected probability level

Describe a Chi Square test

A nonparametric test (ordinal or nominal data) used to compare data in the form of frequency counts occuring in two or more mutually exclusive categories