Statistics Orientation
Darrin Coe, Ph.D.
What?
 Statistics is a mathematical tool for
organizing and understanding information
or data
 Collection of information or data = research
design and methods
 Quantification of information or data = placing
data into analyzable formats.
 Description of information or data = initial global
picture of the data or phenomena
 Analysis of information or data = causality,
prediction, understanding of the data or
phenomena
In many ways statistics is
Simply the study of probability and
odds; possibilities and perhapses
Where? How?
 Statistics is used to study phenomena that
can be operationalized, and quantified:
 Hope and forgiveness -- psychology
 Economic trends -- economics
 Movement of quarks -- physics
 Increases in marijuana yield -- biology
 Probabilities of yield when combining peanut
butter and chocolate – marketing/business
 Inflation change over time -- economics
 Causal effects of isolation on mental illness –
penology
When?
 Statistics and statistical analyses are
used for doing quantitative research
 Used only with data that has been
collected in a quantifiable form and has
been operationalized as a quantity of
some construct.
Who?
 Statistics are generally used by
researchers who are:
 Positivists
 Empiricists
 Realists
 They believe there is an objective,
quantifiable reality that exists beyond
our subjective perceptions.
Data, Data, Data
 types of data
 Nominal – consists of labels, no absolute
zero point, no standardized measure, no
rank ordering (men, women, Floridian,
Dakotan)
 Ordinal – labels, rank ordered, no
standard measure, no absolute zero
point (less happy, happy, more happy)
More Data, Data, Data
 Types of data (continued)
 Interval – rank ordered, standard
measure between intervals, no absolute
zero point (Fahrenheit temperature
scale).
 Ratio - rank ordered, standard measure
between intervals, absolute zero point
(Kelvin temperature scale, number of
eye blinks in 1 minute, number of joints
smoked in 1 hour)
Let’s Get Organized: Summary
Stats
-- used to organize data and provide
overview style picture of a population
or sample.
-- Measures of central tendency
-- Mean, median, mode
-- Measures of variability
-- range, minimum, maximum,
variance, standard deviation.
Standardized Statistics
Z – scores
-- generally used when population
data is known.
T-Scores
-- generally used when population
data is unknown and needs to be
approximated.
Confidence Intervals
-- used to indicate where true scores
fall.
Scores of Relationship
 Correlations – how strongly variables
are related, or how strong one
variable predicts another variable.
-- Pearson’s
-- Used with continuous data
-- Spearman’s
-- used with nominal/ordinal data
So how significant is it?
 Often referred to as the “p-level”
 Generally = .05, .01, or .001
 Also referred to as “alpha”
 Alpha = the probability that you
would get a result given that the null
hypothesis is true in the population.
 P-level is arbitrary and determined by
scientific tradition.
More significance
 A significant result simply means that
a null hypothesis is not true.
 Significance equals “yes” or “no”
 Significance does not equal “proof” or
“truth”
 Significance does not prove the
research hypothesis.
Probability
 The bridge that connects samples with the
populations they are drawn from.
 A proportion consisting of a specific result
over all possible results.
 Given 10 baggies, 3 with MJ and 7 with
ditchweed, the probability of buying a dime bag
of ditchweed is p(dw)=7/10 or 70% probable
you’ll get screwed, so find a new dealer.
Yep, it’s probable.
 Probability guides research sampling
methods.
 Probability is tied to the normal
distribution and allows for the
development of an overall picture of
population by considering the
distribution of a sample.
Inferential Statistics
 Using probability to compare the
means of two or more groups in a
sample to make conclusions about a
given population.
Types of inferential stats
procedures
 Comparing two different groups
 Comparing one group to itself.
 Comparing more than two different
groups
 Comparing one group to itself
multiple times.
 Comparing the effects of multiple
variables on one group.
What?
 Inferential statistics at it’s foundation
uses the standard deviation or the
variance of two or more groups to
compare the means of some outcome
measure or measures.
What??
 This comparison results in a
probability that there is or is not a
statistical difference between the
means. This probability tells us how
much we can infer about the
population from which the sample
was taken.
Final Word
 Statistics is probability.
 Probability is relate to research
design
 Research design is related to validity
 Validity is related to reality.
 Statistics = the probability of reality
Fun huh?

Statistics orientation

  • 1.
  • 2.
    What?  Statistics isa mathematical tool for organizing and understanding information or data  Collection of information or data = research design and methods  Quantification of information or data = placing data into analyzable formats.  Description of information or data = initial global picture of the data or phenomena  Analysis of information or data = causality, prediction, understanding of the data or phenomena
  • 3.
    In many waysstatistics is Simply the study of probability and odds; possibilities and perhapses
  • 4.
    Where? How?  Statisticsis used to study phenomena that can be operationalized, and quantified:  Hope and forgiveness -- psychology  Economic trends -- economics  Movement of quarks -- physics  Increases in marijuana yield -- biology  Probabilities of yield when combining peanut butter and chocolate – marketing/business  Inflation change over time -- economics  Causal effects of isolation on mental illness – penology
  • 5.
    When?  Statistics andstatistical analyses are used for doing quantitative research  Used only with data that has been collected in a quantifiable form and has been operationalized as a quantity of some construct.
  • 6.
    Who?  Statistics aregenerally used by researchers who are:  Positivists  Empiricists  Realists  They believe there is an objective, quantifiable reality that exists beyond our subjective perceptions.
  • 7.
    Data, Data, Data types of data  Nominal – consists of labels, no absolute zero point, no standardized measure, no rank ordering (men, women, Floridian, Dakotan)  Ordinal – labels, rank ordered, no standard measure, no absolute zero point (less happy, happy, more happy)
  • 8.
    More Data, Data,Data  Types of data (continued)  Interval – rank ordered, standard measure between intervals, no absolute zero point (Fahrenheit temperature scale).  Ratio - rank ordered, standard measure between intervals, absolute zero point (Kelvin temperature scale, number of eye blinks in 1 minute, number of joints smoked in 1 hour)
  • 9.
    Let’s Get Organized:Summary Stats -- used to organize data and provide overview style picture of a population or sample. -- Measures of central tendency -- Mean, median, mode -- Measures of variability -- range, minimum, maximum, variance, standard deviation.
  • 10.
    Standardized Statistics Z –scores -- generally used when population data is known. T-Scores -- generally used when population data is unknown and needs to be approximated. Confidence Intervals -- used to indicate where true scores fall.
  • 11.
    Scores of Relationship Correlations – how strongly variables are related, or how strong one variable predicts another variable. -- Pearson’s -- Used with continuous data -- Spearman’s -- used with nominal/ordinal data
  • 12.
    So how significantis it?  Often referred to as the “p-level”  Generally = .05, .01, or .001  Also referred to as “alpha”  Alpha = the probability that you would get a result given that the null hypothesis is true in the population.  P-level is arbitrary and determined by scientific tradition.
  • 13.
    More significance  Asignificant result simply means that a null hypothesis is not true.  Significance equals “yes” or “no”  Significance does not equal “proof” or “truth”  Significance does not prove the research hypothesis.
  • 14.
    Probability  The bridgethat connects samples with the populations they are drawn from.  A proportion consisting of a specific result over all possible results.  Given 10 baggies, 3 with MJ and 7 with ditchweed, the probability of buying a dime bag of ditchweed is p(dw)=7/10 or 70% probable you’ll get screwed, so find a new dealer.
  • 15.
    Yep, it’s probable. Probability guides research sampling methods.  Probability is tied to the normal distribution and allows for the development of an overall picture of population by considering the distribution of a sample.
  • 16.
    Inferential Statistics  Usingprobability to compare the means of two or more groups in a sample to make conclusions about a given population.
  • 17.
    Types of inferentialstats procedures  Comparing two different groups  Comparing one group to itself.  Comparing more than two different groups  Comparing one group to itself multiple times.  Comparing the effects of multiple variables on one group.
  • 18.
    What?  Inferential statisticsat it’s foundation uses the standard deviation or the variance of two or more groups to compare the means of some outcome measure or measures.
  • 19.
    What??  This comparisonresults in a probability that there is or is not a statistical difference between the means. This probability tells us how much we can infer about the population from which the sample was taken.
  • 20.
    Final Word  Statisticsis probability.  Probability is relate to research design  Research design is related to validity  Validity is related to reality.  Statistics = the probability of reality Fun huh?