Biostatistic Flashcards
Nominal
data that is arranged in categories and cannot be arranged in a particular order
Ex: gender, eye color
Ordinal
data arranged in some order, but differences between data cannot be determined or are meaningless
Ex: Stages of PU ( Stage 1- IV)
Interval
data can be arranged in some order and meaningful amount of differences can be determined. There is no true zero.
Ex: Temperature on Fahrenheit
Ratio
data can be arranged in some order and meaningful amount of differences can be determined with a inherent zero starting point.
Ex. Age, monthly income of physical therapists
Symmetric distribution
A distribution where left-hand side is a mirror image of right hand side
positively skewed
A distribution where the scores pile up on the left side and taper off to the right.
negatively skewed
A distribution where the scores pile up on the right side and taper off to the left.
Variability
provides a quantitative measure of the degree to which scores in a distribution are spread out or clustered together
Range
The distance/difference between the largest score and the smallest score
Interquartile range
Interquartile range (IQR) is the distance between Q1 and Q3. Semi-Interquartile range is one half of the IQR.
Variance
is the square of the average distance from the mean.
The SD
approximates the average distance from the mean.
SD = √ variance
Parametric
Used for ratio or interval data
Data must be normally distributed
Data is linearly related
Nonparametric
Used for nominal or ordinal data
Does not rely on normal distribution – used for skewed interval and ratio data
Interval or ratio data – small sample size (n)
Tests relationship
Coefficient correlation is 0-1
1 = perfect association
0 = no association
Parametric Statistics- Test relationship
Dependent variable = outcome (y)
Independent variable = predictors/ factors (x1, x2.. xn)
Pearson product Moment correlation (Pearson’s r) Intraclass Coefficient (ICC) Multiple correlation Linear Regression Multiple Linear Regression
***Correlation describes a relationship between two variables. It does not tell “causality”
Pearson’s r
Reliability of two variables
Association between two variables
Ex: relationship between height (in meters) and weight (in pounds) in adults
ICC
Association between three or more variables
Reliability of repeated measures
Ex: Interrater reliability for strength testing using dynamometer
Multiple correlation
Association between three or more variables
Linear Regression
Studies about prognostic factors
Mathematically a straight line can be described by an equation:
y = b0+ b1x + e1
b0 and b1 have constant values.
b0 (y-intercept of the line): the value of y when x = 0
b1 (slope): the amount of change in y for each 1-unit increase in x
e1 is standard error estimate
Can an outcome “y” be predicted based on factor (s)?
Linear Regression
Can outcome “y” be predicted by one factor “x”? Ex: Can systolic BP be predicted by age
Multiple Linear Regression
Can outcome “y” be predicted by two or more factor “x1, x2..etc.”? Ex: Can systolic BP be predicted by age, weight, LDL levels,…?
Independent
t-test
Between 2 groups
Compare SBP post exercise in 2 groups of elderly: 65-74 & 75-84 yrs
Paired t test
within 2 group test
Compare SBP at baseline and 6 wk post exercise in elderly 65-74 yrs
Analysis of Variance (ANOVA)
Between 2 or more group test
Compare SBP post intervention in 3 groups of elderly: 65-74, 75-84 & >85