You may not have an extinction. You're just missing a lot of information.

193

Set some kind of cutoff on the quality of the record you're going to include in your analysis. Sometimes focusing on the most reliable fraction of the record will be the best way to go.

188