Lectures 1-3 Flashcards
What is biometrics?
biology statistics
what is an entire collection of measurements from all of the organisms that a researcher is interested in
population
What is the numerical summary of the population
Parameter, can RARELY be calculated, very large = not feasible to obtain
What is a parameter estimated by?
a statistic
How is a parameter represented?
greek letters
Examples of parameters
mean and standard deviation
What is the subset of a population
Sample
Samples need to be
random and interspersed
random (sampling)
if every individual in the population has an equal chance of being represented in the sample
interspersed (sampling)
representative of entire populations/areas (i.e.: few fish from deep end and few from shallow end)
What are samples that are not random/interspersed, do not accurately reflect the population of interest
Biased samples
What happens if randomly-selected samples are not interspersed?
it is not representative, a grid is usually used
What is the numerical summary of a SAMPLE
Statistics
What are estimates of population parameters?
statistics
What is the difference between the sample statistic and the population parameter
sampling error
how is objectivity achieved?
by statistical conclusions having a bias in probability
which parts of the scientific method does biometrics focus on?
analysis of the results and drawing a conclusion based on your statistical analysis
What are numerical facts, pieces of info
Data (datum = one piece of data)
the relationship between statistics and data
both are numerical info but statistics is more used
What are objects defined by a set of data
individuals
What are characteristics of the individuals (i.e.: weight, height, hair color)
variables
relationship between data, individuals, and variables
calculate statistics and then used for parameter
types of variables (data) scales
nominal, ordinal, ratio, interval
What type of data are qualitative (description/categories)
nominal scale data
What type of data are quantitative, ranked data due to unequal increments between successive values | non-parametric tests
ordinal scale data
Which type of data is quantitative, equal increments between successive values (measured values), NO biologically meaningful zero
interval scale data (i.e.: temperature)
Which type of data is quantitative, equal increments between successive measured values, HAS biologically meaningful zero
ratio scale data (i.e.: lbs, concentration, ft.)
what is important for employing appropriate statistic
making the distinction between these types of data
two types of statistics
inferential and descriptive
What are estimations of population parameters
inferential statistics
What are the numeral summary of statistics
descriptive statistics
two types of descriptive statistics
measures of: central tendency and dispersion
what are parts of measures of central tendency
mean, median, mode
what appropriate measure of central tendency is used for ratio-interval scale data, most common for interval
mean (average)
what is mean of several means
grand (weighted) mean
What measure of central tendency is used for ordinal-scale data
median, middle value in an ordered set of data
what measure of central tendency is used for nominal-scale data
mode, number that occurs most frequently in the data set
what is measures of dispersion
variability used for ratio-interval scale data
what is range?
difference between the largest and smallest value
in which type of data is range used for?
ordinal-scale
what are diversity indices
used for nominal-scale data
What shows measured values (data) arranged from smallest to largest on the x-axis, and the frequency that these values occur on the y-axis
frequency distributions
what are bar graphs used in?
nominal scale data
how can frequency distributions be represented?
histograms or probability curves
what are frequency distributions useful for?
determining the probability of obtaining certain values
What is the frequency distribution that most complex variables (biological variables) follow
normal distributions
what are properties of a normal distribution
symmetrical at the mean, bell-shape curved
what defines the location/center of a normal distribution
the mean
what defines the shape/spread of a normal distribution
the standard deviation
how many normal distributions are there?
infinite due to infinite mean and standard deviation combinations
skewness
asymmetry of a distribution curve
what skew represents a higher frequency of larger values
right skew
what skew represents higher frequency of lower values
left skew
what describes higher “peakness” in a frequency distribution curve
kurtosis
what is leptokurtosis
all of the values occur right at the mean – all other values are basically nonexistent
what is platykurtosis
the greatest frequency on all the values – basically all the values have the same frequency
purpose of proportions of a normal distribution
to determine the probability of obtaining certain values
what has a mean of 0, and a standard deviation of 1, values are standardized by converting them into Z-scores
the standard normal distribution
what are data distributions
the frequency distribution of raw data
What is the standard deviation of several means values
standard error of the mean (SE)
What is the frequency distribution of the means of the raw data (many means)
sampling distributions
what is the best estimate of what we could get if we could get samples over and over
standard error of a single sample
(central limit theorem) what does it mean if the samples of any size are taken from a normally distributed population
the means of these samples will be normally distributed as well
(central limit theorem) what will samples taken from any distribution have?
means that approach normal distributions as the sample sizes increase
(central limit theorem) what is the relationship between the standard error and sample size
standard error will decrease as the sample sizes increase