An ozone trend analysis that compensates for sampling biases is applied to sparsely sampled occultation data sets. International assessments have noted deficiencies in past trend analyses and this work addresses those sources of uncertainty. The nonuniform sampling patterns in data sets and drifts between data sets can affect derived recovery trends by up to 2 % decade⁻¹. The limitations inherent to all techniques are also described and a potential path forward towards resolution is presented.