2000
DOI: 10.1007/s002549900081
|View full text |Cite
|
Sign up to set email alerts
|

Normal and lognormal data distribution in geochemistry: death of a myth. Consequences for the statistical treatment of geochemical and environmental data

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
1
1
1
1

Citation Types

9
247
0
4

Year Published

2014
2014
2018
2018

Publication Types

Select...
10

Relationship

0
10

Authors

Journals

citations
Cited by 562 publications
(260 citation statements)
references
References 0 publications
9
247
0
4
Order By: Relevance
“…The concept of transforming data in order to reduce asymmetry of an empirical density distribution has been widely suggested in geochemical data analysis (Campell, 1982, andSinclair, 1983). Log-transformation of data is the most common numerical transformation used for geochemical or environmental dataset (Reimann and Filzmoser, 2000). However, in this study, the square root-transformed uni-elemental data subset better adjusted to the normal distribution compared to the log-transformation (data not shown).…”
Section: Outlier Detection and Background Determinationmentioning
confidence: 61%
“…The concept of transforming data in order to reduce asymmetry of an empirical density distribution has been widely suggested in geochemical data analysis (Campell, 1982, andSinclair, 1983). Log-transformation of data is the most common numerical transformation used for geochemical or environmental dataset (Reimann and Filzmoser, 2000). However, in this study, the square root-transformed uni-elemental data subset better adjusted to the normal distribution compared to the log-transformation (data not shown).…”
Section: Outlier Detection and Background Determinationmentioning
confidence: 61%
“…Log transformations failed to normalize data, so the Mann-Whitney nonparametric test for comparison of medians was used to assess differences between groups. For ferulic acid and caffeic acid, values that were below the range of detection were assigned a value of one half the minimum value [22].…”
Section: Discussionmentioning
confidence: 99%
“…The performance of statistical analysis can be affected dramatically by data outliers (Reimann and Filzmoser 2000). We excluded two outliers from the raw dataset prior to statistical analysis processes, which had PAH concentrations 19 times higher than the standard deviations of the rest of samples apparently caused by local point sources rather than regional factors.…”
Section: Data Preparation and Statistical Analysismentioning
confidence: 99%