Card Set Information

2014-10-20 16:58:36
yup stats
Show Answers:

    • author "amcmullen"
    • tags "Vocab"
    • folders "Stats"
    • description "statistics"
    • fileName "stats2"
    • Population
    • a collection of entities we want to study
  1. Sample
    a subset of elements from a population
  2. Random Sample
    From a quantitative perspective, each element has an equal opportunity of being selected
  3. External Validity
    A study has external validity when the findings of the sample can be generalized to the population of interest. When selected accurately the sample represents the population from which it was drawn
  4. Random Selection
    • From a quantitative perspective, a process whereby each element of a population of interest has an equal opportunity of being selected
    • (relates to external validity)
  5. Random Sample
    • A subset of the population whose information can be used to describe the entire population with a measured level of confidence
    • (relates to external validity)
  6. Internal Validity
    When a study measures what it purports to measure
  7. Random Assignment
    • Assigning subjects (usually from a sample) randomly to conditions to avoid systematic groupings that may hinder clear interpretation of the results
    • (relates to internal validity)
  8. Variable
    property of an entity of interest that can take on many values
  9. Discrete Variable
    variable that has no intermediate values between scale values
  10. Continuous Variable
    variable that has an infinite number of values between scale values
  11. Measurement
    assigning value to a property of an entity, assigning value to a variable
  12. Measurement Data or Quantitative Data
    • data that is obtained through some form of measurement
    • values indicate amount (more/less) or degree of a variable
    • values are continuous in their nature
  13. Categorical, Frequency, or Qualitative Data
    • data that is obtained via labeling or categorization
    • values are discreet
  14. Independent Variable
    the variable manipulated by the experimenter, but conceptually not related to another variable
  15. Dependent Variable
    the variable not under the experimenter's control and depends on the independent variable
  16. Descriptive Statistics
    • a number that is describes a specific characteristic of a set of data
    • mean
    • median
    • mode
  17. Inferential Statistics
    • statistics about a sample that is used to make inferences (educated guesses) about a population
    • different populations
    • making inferences about where the group is
  18. Parameter
    a measure referring to or deriving from the population
  19. Statistic
    a similar measure as a parameter that derives from a sample
  20. Nominal Scale
    • the value of a variable that has a nominal scale is just a label or a name¬†
    • (almost all values operate on a nominal scale)
  21. Ordinal Scale
    • Similar to the nominal scale but the order of the labels is important
    • The differences between the labels do not have to be equal
    • Includes nominal
    • e.g. - class rank
  22. Interval Scale
    • In addition to having the same properties of a variable within an ordinal scale, the difference between units of measurement is similar
    • Includes nominal and ordinal
    • e.g. - temperature (Celsius and Farenheit)
  23. Ratio Scale
    • In addition to have the same properties of a variable with an interval scale, this type of variable also has a true 0 point
    • Includes all the values of nominal, ordinal, and interval
    • e.g. Kelvin temperature
  24. Central Tendency
    • Measures refer to a "middle" or "typical" value of the data we just collected, including:
    • Mode
    • Median
    • Mean
  25. Mode
    • The value that occurs the most often in a distribution
    • The value has a greater likelihood of being selected from a distribution of scores
    • Appropriate for use with all scale variables
    • Not as informative as other measures of central tendency since it merely uses one score from the whole distribution
    • Not affected by extreme scores
    • If the scores are too far apart, they are not averaged and the distribution is called bimodal
  26. Median
    • 50th percentile - middle of distribution - half above and half below
    • Median Value - middle number - add up middle scores and divide by 2
    • Median Location - (N+1)/2
    • Not as useful as other measures of central tendency since it merely represents a value that may not even exist in the distribution
    • Not affected by extreme scores
    • Appropriate for ordinal scale variables, and interval/ratio scales variables when the distribution is skewed
  27. Mean
    • The average value from a collection of scores
    • Equal to the sum of the scores divided by the number of scores
    • Appropriate for variables with interval/ratio scales
    • Influenced by extreme scores and requires
    • Easily manipulated algebraically
    • Stable across multiple samples of a population, making it a good estimate of central tendency for the population
    • Population¬†
    • Sample x-bar =¬†
  28. Measures of Variability
    • Focus on the dispersion of values around the mode, median, or the mean
    • Include - range, average deviation, variance, standard deviation
  29. Range
    • Reflects the distance between the lower and highest value
    • Highly influenced by extreme "outlier" scores that are not "representative" in the true population
    • Interquartile range removes the top and bottom 25% of scores, eliminating extreme values from consideration
    • Reduced data is a "trimmed sample"
  30. Average Deviation
    • The difference between each score and the mean of all scores in a sample can be summed and divided by the number of scores in that sample
    • The sum of the deviations from the mean sum up to 0
    • Not a useful measure of dispersion
  31. The Mean Absolute Deviation (M.A.D.)
    • Takes the absolute value between each score and the mean of all the scores
    • The deviations no longer sum up to 0
  32. Variance
    • An alternative to the M.A.D. is the variance
    • Population Variance = 2
    • Sample Variance = s-hat2
    • Instead of taking the absolute value of the dispersion, square it
    • The inflated values resulting from the squaring of the differences allow us to detect small dispersion that would not be as visible using the M.A.D.
    • The squared (INFLATED) values do not correspond to those in the data set. The standard deviation (square root of the variance) does
    • Highly influenced by extreme values