Contents

VARIABLES

BETWEEN

O

M

I

ONEINPUTVARIABLE:

*

*

*

C

C

C

USED

TO

TEST

IF

TWO

VARIANCES

ARE

EQUAL.

F-RATIO

EXAMPLE-ICES

M

F-RATIO

EXAMPLE-ICES

F-RATIO

EXAMPLE-ICES

C
COMPARING VARIABILITY WITHMULTIPLE LEVELS

  BARTLETT'STEST:

  Overhead 3-6

C

  COCHRAN'STEST:

  SAMPLE SIZES MUST BE (APPROXIMATELY) EQUAL

CRITICALVALUESFORCOCHRAN'STEST
  k
  2
  3
  6
  7
  8

CRITICALVALUESFORCOCHRAN'STEST
  k
  9
  10
  37
  145

DIFFERENCE

BETWEEN

MEANS

OF

TWOPOPULATIONS

OR

CRITICALVALUESOFT
  .10
  .005
  .0005

CRITICALVALUESOFT

  (Cont'd)
  .10
  .005
  .0005

CRITICALVALUESOFT

  (Cont'd)
  .10
  .005
  .0005

N-C

ANOVASTARTSWITHA "MODEL"

ANOVASTARTSWITHA "MODEL"

  (Cont'd)

ANOVA

ANOVA

  (Cont'd)

ANOVA

  (Cont'd)

C

ANOVA- REVIEW

TWOPROPORTIONS
  P


  P

  P

  =
  =
  =
  =
  Proportion bad in group 1
  Proportion bad in group 2
  Proportion bad overall :
  Total tested in group 1

  H

  P

  P

  Total tested

  Z

  =
  Requirement :
  Total bad
  5, total good
  5

CORRELATIONAND

CAUSATION

"THEMEANING

OF

R

  Y

  =

  x

  . . . ., x
  +
  +
  . . . .
  R²
  The proportion of the variance Y

CRITICALVALUESOFTHE

PEARSON
  .10
  .05
  .02
  .01

CRITICALVALUESOFTHE

PEARSON
  .10
  .05
  .02
  .01

CRITICALVALUESOFTHE

PEARSON
  .10
  .05
  .02
  .01

CORRELATION

COEFFICIENT

PREDICTIONAND LINEAR REGRESSION

(B
  CAP-5000

EXAMPLE