|
Statistical tools available in QP3.0
1.
To describe parasitic infection of a single sample
of hosts:
• Descriptive statistics (N
hosts, prevalence, mean & median intensity,
variance/mean),
• Alternatives to calculate
an exact confidence interval for the prevalence
∙ Clopper - Pearson method
(this is more traditional),
∙ Sterne or Wald method (this
is more advanced, see Reiczigel 2003),
• Bootstrap (BCa) confidence
interval for the mean intensity,
• Exact confidence interval
for the median intensity,
• Bootstrap (BCa) confidence
interval for the mean abundance,
• Bootstrap (BCa) confidence
interval for the mean crowding (see Reiczigel et al.
2005a),
• Aggregation indices
(variance/mean, index of discrepancy, and k of the
negative binomial),
2. To compare parasitic infections between two
samples of hosts:
• Descriptive statistics (N
hosts, prevalence, mean & median intensity,
variance/mean),
• Alternatives to compare
prevalences:
∙ Chi-square Test (this is a
more traditional way),
∙ Fisher's Exact Test
(this is a better alternative),
∙ Unconditional Test
(most advanced, see Reiczigel et al. 2008.)
NEW (02. 01. 2010.)
• Bootstrap t-Test to compare
Mean Intensities,
• Mood’s Median Test to
compare Median Intensities,
• Stochastic equality of
intensity distributions (see Reiczigel et al.
2005b),
• Bootstrap t-Test to compare
Mean Abundances,
• Comparison of Mean Crowding
(see Reiczigel et al. 2005a),
3. To compare parasitic infections among samples of
hosts:
• Descriptive statistics (N
hosts, prevalence, mean & median intensity,
variance/mean),
• Chi-square Test to compare
Prevalences (this is a more traditional way),
• Fisher's Exact Test to
compare Prevalences (this is a better alternative),
• Mood’s Median Test to
compare Median Intensities.
|
