Shared Flashcard Set

Details

Biometrics Midterm 1
concepts
62
Biology
Undergraduate 3
02/08/2015

Additional Biology Flashcards

 


 

Cards

Term
biometrics
Definition
to measure life using statistics appied to life
Term
population
Definition
entire collection of measurements from all the organisms the reasearcher's interested in (total)
Term
parameter
Definition
numerical feature of the population (mean)
Term
sample
Definition
subset of population
Term
random samples
Definition
every single individual in population has a chance
Term
interspersed samples
Definition
representative of population
Term
biased
Definition
does not accurately reflect the population of interest
Term
statistic
Definition
numerical feature of the sample
Term
sampling error
Definition
difference between sample statistic and the population parameter
Term
scientific method
Definition

1) observations/questions

2) hypothesis

3) experiment

4) results

5) analyze results with statistics

6) conclusion

Term
individual
Definition
- objects described by a set of data
Term
variables
Definition
characteristics of individuals
Term
nominal-scale data
Definition

qualitative (descriptive) data

ex. gender, eye color

Term
ordinal-scale data
Definition
quantitative, unequal intervals, ranked values
Term
interval-scale data
Definition
quantitative, equal increments, no meaningful zero value
Term
ratio-scale data
Definition
quantitative, equal intervals, meaningful zero value
Term
inferential statistics
Definition
estimates of population parameters
Term
numerical statistics
Definition
numerical feature of a sample
Term
mean
Definition
central tendency fo ratio-interval scale data
Term
grand (weighted) mean
Definition
mean of means
Term
median
Definition
measure of central tendency used for ordinal-scale data (ranked) data
Term
mode
Definition
measure of central tendency used for nominal-scale data
Term
dispersion
Definition
(variability) sum of squares, variance, standard deviation, coefficient of variation (ratio-interval scale)
Term
range
Definition
measure of dispersion for ordinal-scale (difference between highest and lowest value)
Term
diversity indices
Definition
measures of dispersion for nominal-scale data (qualitative)
Term
frequency distributions
Definition

determines the probability of obtaining certain values 

(histogram, probability density curve)

Term
normal distributions
Definition

greatest frequency of occurrence is at mean

lowest frequency of occurence is at extremes

distribution is symmetrical about the mean

Term
right skewness
Definition
asymmetrical, greater frquency of large values
Term
left skewness
Definition
asymmetrical, greater frequency of extreme small values
Term
kurtosis
Definition
peakedness, symmetrical
Term
leptokurtosis
Definition
all values are around mean
Term
platykurtosis
Definition
plateau, not peaked, greatest frequency are all values
Term
standardized normal distribution
Definition

mean - 0

standard deviation - 1

values converted to Z-scores

Term
data distributions
Definition
frequency distribution of raw data
Term
sampling distributions
Definition
frequency distribution of statistics (not raw data)
Term
standard error
Definition

- standard deviation of many means

- single sample 

- estimate of sampling distribution if multiple samples taken

Term
central limit theorem
Definition

- means of samples normally distributed as sample size increases

- standard deviation will decrease as sample size increases

Term
null hypotheses
Definition
no difference or no relationship
Term
alternative hypotheses
Definition
there is a difference or relationship
Term
there is a difference because
Definition

1) populations are different

2) sampling error

Term
objectivity
Definition
achieved by using probabilities
Term
P-value
Definition
probability of obtaining your sample data and statistics from a population in which null hypothesis is true
Term
type I error
Definition
rejecting null hypothesis when it is true
Term
type II error
Definition
accepting null hypothesis when it is false
Term
goodness-of-fit test
Definition

- one nominal-scale variable, compare frequency distribution to a priori ratio
-  DF = k -1

- R&B = n/k > 2

Term
purpose of continuity correction
Definition
reduce risk of getting large critical value and reduce risk of type I error
Term
contingency table
Definition

- two nominal-scale variables

- frequency of occurrence of categories is independent of frequency of occurence of other variable

- DF = (r-1)(c-1)

- R&B = n/(r x c) > 6

- continuity correction (2x2)

Term
one sample t-test
Definition

- probability that sample of ratio-interval scale taken from a population with pre-determined mean 

- mean = c (a priori constant)
- DF = n-1
two tailed: no difference

one tailed: direction of difference 

Term
confidence intervals
Definition
mean of population lies between two values
Term
independent samples
Definition
if values in one sample are not related in any particular way to values in other sample
Term
paired samples
Definition
if each value in one sample is associated with one particular value in the other sample
Term
assumptions for two sample testing
Definition
RSNDP (normal distribution)
- D'Agostino Pearson K^2 test for normality
Homoscedastic (equal variances)
Term
types of data transformation
Definition
logarithmic transformation, square root transformation, arcsine square root transformation
Term
nonparametric test
Definition
assumptions are not met, ordinal-scale data. do not test for differences in the parameter (mean)
Term
two sample t-test
Definition
- tests for differences in the means of two independent samples of ratio-interval scale data
- DF= (n1-1)(n2-1)
- normal distributions, equal variances
- nonparametric alternative: Mann-Whitney U test
Term
Mann-Whitney U test
Definition
tests for differences in the central tendency of two independent samples of ratio-interval scale data and ordinal-scale data (ranking) when assumptions of normality/homoscedasticity are not met
- DF = n
Term
outliers
Definition
data points that are much more extreme than all of the other data in the sample
Term
paired sample t-test
Definition
- differences in the means of two paired samples of ratio-interval scale data, comparing difference between two means to zero
- DF = n(d) - 1
Term
wilcoxon paired sample test
Definition
- differences in the central tendency of two paired samples of ratio-interval scale data when the normality assumption is not met
- ordinal-scale data
- calculated test stat uses n (sample size)
Term
logarithmic transformation
Definition
x' = log x
consistent (most common)
Term
square root transformation
Definition
x' = rad x
whole number counts
Term
arcsine square root transformation
Definition
x' = arcsin rad x
when data are proportions or percents
Supporting users have an ad free experience!