Docstoc

Statistics

Document Sample
Statistics Powered By Docstoc
					                 Statistics
Variable
     Anything that can have more than one
     value
           ie. Height, weight, sex, IQ, video games


Constant
    Anything that remains the same
           ie. Study on females (females remains
               constant)
          Statistics cont.
Utilizes the organization of raw data that has
already been collected
    Expressed in different ways:
          1. Histogram (Bar graph)
               x axis (abscissa) - horizontal
               y axis (ordinate) - vertical
          2. Pie Chart
          3. Polygon (Line graph)
 Which class performed better?

            Class A                                  Class B

90                                        90

                                          75

85                                        60

                                          45
                              % on test                                % on test
80                                        30

                                          15

75                                        0
     Ss 1   Ss2   Ss3   Ss4                    Ss1   Ss2   Ss3   Ss4
       Measurement Scales
Nominal Scale:
    numbers are used to name or categorize
    (ie. Sports uniforms, gender,driver’s licence)


Ordinal Scale:
    numbers represent serial position
    (ie. Finish in a race, rankings, level of
    aggression)
    Measurement Scales cont.
Interval Scale:
     consistent units if measurement and equal
     spacing between units (no true zero point)
          ie. Celcius

Ratio Scale:
     Same as interval, but has absolute zero point
          ie. Weight, height, time, distance
           Types of Statistics
Descriptive Statistics
     - A statistic to describe behaviour
        - ie. SAT scores


Inferential Statistics
     - A statistic that explains behaviour
        - Bandura’s Bobo doll experiment on aggression
     - example: t-test or ANOVA (analysis of variance)
  Measures of Central Tendencies
A single score that represents a whole
set of scores


Mode: Most frequently occurring score

Median: Middle score (half above and below)

Mean: Arithmetic average (most widely used)
           Skewed statistics
When the data is not distributed evenly,
atypical scores can mislead the data


Always be aware of the measure of central
  tendency that people are reporting
           Income data
$20,000
$22,000
$23,000
$28,000
              What is the Mode?
$34,000
$38,000
$41,000       What is the Median?
$42,000
$44,000
$85,000
$90,000
              What is the Mean?
$210,000
$210,000
    Normal Curve (Bell Curve)
Theoretical or hypothetical frequency of
 scores

68% between 1 and -1 SD
95% between 2 and -2 SD
99% between 3 and -3 SD

Examples: height, weight, IQ
                      Variability
The amount of variability between scores tells us the reliability
of the data
      The lower the variability the more reliable the results


Basketball player example


Range: The gap between the lowest and the highest score

        Crude estimate of variability based on extreme scores
             Standard Deviation
- Standard measurement of how the scores in a
   distribution deviate from the mean

- Better way to assess level of variability between
  scores

- Represented by lower case s
- ie. UBC vs. Kwantlen entrance marks
              who has the smaller SD?
  When is an observed difference
             reliable?
1. Representative vs. Biased Sample
     Is it random and large enough?

2. Less variable observations are more
  reliable

3. More cases are better
    ie. Assessing which school is better?
         Statistical Significance
- Statistical statement of how likely it is that an
  obtained result occurred by chance

- p<.05
   - “beyond a reasonable doubt”

   - indicates the likelihood the result will happen by
     chance
          - Does not indicate IMPORTANCE of result
          - ie. IQ between first born and second born
         Reliability & Validitiy
 Reliability is the extent to which an
  experiment, test, or any measuring
  procedure yields the same result on
  repeated trials.
 Validity refers to the degree to which a
  study accurately reflects or assesses the
  specific concept that the researcher is
  attempting to measure.
         Reliability & Validity
 While reliability is concerned with the
  accuracy of the actual measuring
  instrument or procedure, validity is
  concerned with the study's success at
  measuring what the researchers set out to
  measure.
Reliability & Validity
      Correlation Coeffiecient
Measures relationship between two
 variables
    - data must be interval or ratio

Pearson Product-Moment Correlation is
 represented by small case r
                remember…
If anything is paid attention to in our schools,
   colleges, and universities, thinking must be it.
   Unfortunately, thinking can be lazy. It can be
   sloppy….It can be fooled, misled, bullied...
   Students possess great untrained and untapped
   capacities for logical thinking, critical analysis,
   and inquiry, but these are capacities that are not
   spontaneous: They grow of wide instruction,
   experience, encouragement, correction, and
   constant use.
               Experiment
Information:
  The Nestle Inc. authoritatively states that
  Smarties are 25% pink, 20% green, 15%
  brown,15% yellow, 15% orange, and 10%
  purple. This distribution is important for
  “quality control”.

Is this true? Are they that accurate?
                   Sample
Three random boxes of mini Smarties. All boxes
  are still originally sealed and have not been
  tampered with.

Design a data sheet and complete your data
 collection. (Ensure that you complete your data
 before premature subject mortality occurs!)

Convert raw data into percentages
           Hypothesis


Develop a hypothesis based on the
distribution of the data you collected
      Proving your hypothesis
Join up with two other classmates and
  integrate your data.

Is your hypothesis still right?

Do you have a new hypothesis?

				
DOCUMENT INFO