2. Statistical Cross Dating - emboylen/Foxfish_chronology GitHub Wiki

Cross-dating is the process of validating the correct year is assigned to the correct increment. It is based on the assumption that your sample is synchronous; wide bands should fit with side bands, and narrow with narrow and so on. This is essentially how cross-dating is carried out - it plots the data and looks for increments which don't seem to be fitting the trend of the other increments. Cross-dating is important as errors are common in construction chronologies, increments can be skipped, counted twice or assigned the incorrect year.

Cross-dating can be done either visually - with "skeleton plots" or statistically with a chronology program. Many studies use both visual and statistical cross-dating techniques. The conventional program used in dendrochronology and aquatic biochronologies is COFECHA (which is a free software found here). Visual-cross-dating has been scrutinized as being not very useful, and not achieving anything different than statistical cross-dating.

Once cross-dating (visual or statistical) has taken place, the highlighted outliers are inspected in the otolith or the data. This is a dangerous practice: the data may be "forced" into showing synchrony, when synchrony may not be present. One must always be careful to only change an increment if an error is present.

For my honours thesis, I will be using a combination of visual and statistical cross-dating only..

The program I will be using for cross-dating is dplR. dplR is a dendrochronology package in R. The cross-dating techniques have not been utilized in many biochronolohuy studies, as COFECHA is the norm.
From dplR:

"As with any dating enterprise, statistical crossdating is merely a tool and users should always rely on the wood to accurately date tree-ring data. The Dendrochronology Program Library in R (dplR) is a package for den- drochronologists to handle data processing and analysis."