Summary Overview of statistics used in Marketing Research Methods

Type of Statistic Used in Analysis Description Critical value Action
P-value All analyses <0.05 Reject null
hypothesis.
KMO Factor Analysis Kaiser-Meyer- 0.5 Drop variable
Olkin with lowest
KMO if lower
than 0.5.
Communality Factor Analysis Percentage of Should be larger If larger than 1
variance of a than 0.4, but or smaller than
given variable smaller than 1. 0.4, get rid of it.
explained by all
extracted
factors.
Eigenvalue Factor Analysis How much >1 Get rid of it if
variance is smaller than 1.
explained by a
factor.
Cumulative total Factor Analysis How much of >60% Get rid of
explained variance total variance is variables when
explained by 60% is achieved.
cumulative
factors.
Factor variance Factor Analysis How much >5% If smaller than
variance is 5%, get rid of it.
explained by a
single factor.
Rotated loading Factor Analysis Strong >0.5 Delete item if
connection to lower than 0.5.
one dimension
(not more than
one dimension,
otherwise this
might be an
indication that
your factors are
not good).
Cronbach’s Alpha Cluster Analysis Measures >0.6 Delete item with
internal CA lower than
consistency of 0.6, but if you
factor. have only two
items left, do
not delete any.
F-test ANOVA, ANCOVA, Measures Should be as Reject
Multiple significance. large as possible, hypothesis if p
Regression can be value is smaller
transformed in p than 0.05 (=
value. significance).