statistics
Home > Flashcards > Print Preview
The flashcards below were created by user
kimkrak
on
FreezingBlue Flashcards. What would you like to do?

statistics deals with
collection, classification, analysis, interpretation ( of numeral facts or data)

nominal
 numbers ( simple count number of times something has occurred)
 simplest data

ordinal
 order
 these numbers cannot be averaged
 who came 1st 2nd 3rd
 intervals between each are unknown

interval and ratio
 measurement
 intervals known and equal
 ratio has absolute zero point

measure of central tendency

mode
 occurs most
 not influenced by extreme scores

median
 number in middle
 not influenced by extreme scores
 ideal for ordinal

mean ( average )
 DISTORTED BY EXTREME VALUES
 interval or ratio

descriptive statistics
number that summarize set of research data

standard deviation
average amount of scores deviated from the mean

normal distribution
 represents data about how certain human characteristics are dispersed in the population
 the mean median and mode are all in the same place

68% of the scores fall within ____ standard deviationĀ of the mean
95
99

skewed
 extreme values (outliers)
 the MEAN (average) is pulled toward the outliers

correlation=
relationship

scatter plots
best way to examine data on two variables at once

positive correlation
one predicts the other

negative
 negative slope
 presence of one predicts the absence of the other

no correlation
no relationship between variables

correlation coefficent
 strength of association of two variables
 +1 is strongest possible
 0 is no correlation
 1 is strongest possible negative
+1 & 1 are equal in strength
lowercase r is the symbol

correlation does not imply
causation

propability =
# of ways an event can occur/ # of possible outcomes

5% significance is
sufficent

null hypothesis
 opposite of hypothsis being tested
 researcher tries to reject it

type one error
 to lenient
 50 to 50
 or 1 in 3
 researcher may be eager to reject null hypotheis even if it is true

type 2 error
 1 in a 100
 1 in a 1000
 too strict
 researcher compelled to accept null hypothesis

meta anaylasis
pool data from multiple studies until they arrive at an answer